DATA WAREHOUSING AND DATA MINING

COURSE OBJECTIVE: Data Warehousing and Data Mining are advanced recent developments in database technology which aim to address the problem of extracting information from the overwhelmingly large amounts of data which modern societies are capable of amassing. Data warehousing focuses on supporting the analysis of data in a multidimensional way. Data mining focuses on inducing compressed representations of data in the form of descriptive and predictive models. This course gives students a good overview of the ideas and the techniques which are

behind recent developments in the data warehousing. It also makes students understand On Line Analytical Processing (OLAP) , create data models, work on data mining query languages, conceptual design methodologies, and storage techniques. It also helps to identify and develop useful algorithms to discover useful knowledge out of tremendous data volumes. Also to determine in what application areas can data mining be applied. this course also speaks about
research oriented topics like web mining, text mining are introduced for discussing the scope of research in this subject. At the end of the course the student will understand the difference between a data warehouse and a database and also will be capable to implement the mining algorithms practically. SYLLABUS:

III Year IT II Semester DATA WAREHOUSING AND MINING UNIT-I Introduction: Fundamentals of data mining, Data Mining Functionalities, Classification of Data Mining systems, Major issues in Data Mining, Data Warehouse and OLAP Technology for Data Mining Data Warehouse, Multidimensional Data Model, Data Warehouse Architecture, Data Warehouse Implementation, Further Development of Data Cube Technology, From Data Warehousing to Data Mining, Data Preprocessing: Needs Preprocessing the Data, Data Cleaning, Data Integration and Transformation, Data Reduction, Discretization and Concept Hierarchy Generation Online Data Storage. UNIT-II Data Warehouse and OLAP Technology for Data Mining: Data Warehouse, Multidimensional Data Model, Data Warehouse Architecture, Data Warehouse implementation, Further Development of Data cube technology, from data warehousing to data mining. Data cube computation and data generalization: efficient methods for data cube computation, Further Development of Data cube and OLAP Technology, Attribute Oriented induction. UNIT-III Mining frequent patterns , association and correlation: Basic concepts, Efficient and scalable frequent itemset mining methods, Mining various kinds of association rules, From association mining to correlation analysis, Constraint based association mining. UNIT IV Classification and prediction: Issues regarding classification and prediction, Classification by Decision tree introduction, Bayesian Classification, Rule-based classification, Classification of Back propagation, Support vector machines, Associative Classification, Lazy learners, Other classification methods, Prediction, Accuracy and error measures, Evaluating the accuracy of classifier or predictor, Ensemble methods- increasing. UNIT V

Cluster analysis introduction: Types of Data in cluster analysis, A Categorization of major clustering methods, partitioning methods, Hierarchical methods, Density Based methods, Grid based methods and model based clustering methods, Clustering high dimensional data, Constraint based cluster analysis, Outlier analysis. UNIT VI Mining streams, time series and sequence data: Mining data streams, Mining time series and sequence data , Mining sequence patterns in transactional databases, Mining sequence patterns in biological data , Graph mining, Social network analysis, Multirelational data mining. UNIT VII Mining objects, spatial multimedia , text, and web data: Multidimensional analysis and descriptive mining of complex data objects, Spatial data mining, Multimedia data mining , Text Data mining, Mining the world wide web. UNIT VIII Application and trends in data mining: Data mining application, Data mining system products and research prototypes, Additional themes on data mining , Social impacts of data mining.
SUGGESTED BOOKS: (i) TEXT BOOKS: T1. Jiawei Han & Micheline Kamber. Data Mining Concepts and Techniques, Kaufman Publishers

T2. Tan, Pang-Ning and other Introduction to Data Mining Pearson Education, 2006.
(ii) REFERENCES: 1. Arun K Pujari, Data Mining Techniques, University Press. 2. Sam Anahory & Dennis Murray, Data Warehousing in the Real World, Pearson Edn Asia. 3. K.P.Soman, Insight Into Data Mining, PHI 2008. 4. Paulraj Ponnaiah, Data Warehousing Fundamentals, Wiley Student Edition. 5. Ralph Kimball, The Data Warehouse Life cycle Tool kit, Wiley Student Edition.

SESSION PLAN: Subject : Data Warehousing and Mining

Year/ Sem: IT III/II

Topics in each unit as per JNTU Syllabus Introduction Fundamentals of data mining Data mining functionalities Classification of Data mining system, Data mining task primitives Integration of data mining system with a database or a data warehouse system, Major issues in data mining Data Preprocessing: Need preprocessing the data Data Cleaning Data Integration and Transformation Data Reduction Discretization and concept hierarchy generation

Lectu re No 1 2 3 4

Modules / Sub modules for each topic UNIT-I Overview of all units and course delivery plan Data mining fundamentals

Text Books / Reference Books Syllabus book T1: 1.1-1.3, T2:1.1 R4:pg 2-7 T1: 1.4.1- 1.4.6 T2:2.1,2.2 T1: 1.6 , 1.7

What kinds of patterns can be mined Data mining as a confluence of multiple disciplines, Primitives for specifying a data mining task
Integration of data mining system with a database or a data warehouse system,

T1: 1.8,1.9

6 7 8 9 10 11

Need preprocessing the data Missing values, Noisy Data, Inconsistent Data Data Integration, Data Transformation Data Cube Aggregation, Dimensionality Reduction Data Compression, Numerosity Reduction For Numeric data, For categorical data Revision of Unit-I UNIT II What is datawarehouse and its importance

T1: 2.1,T2:2.3, R3:5.1,2 T1: 2.3.1-3, T1: 2.4.1-2.4.2 T1: 2.5.1-2.6.2 T2:2.3.3,4 T1: 2.6.1-,2 R3:5.3

Data warehouse and OLAP technology for data mining: Data Warehouse Multidimensional data model

12

13

From tables and spread sheets to data cubes, Stars, snowflakes and fact constellations schemas for multidimensional databases Steps for design and construction of data warehouse A three tier data warehouse architecture Types of OLAP servers Efficient computation of data cubes Indexing OLAP data, Meta data Repository, Data warehouse Back-end tools Further Development of Data cube Technology Data warehouse usage From OLAP to OLAM Different kinds of cubes, For full cube BUC, Star cubing for fast high dimensional OLAP, Cubes with iceberg conditions

T1:3.1, R2:1.2 R3:2.5.5, R4:Pg12, R5: pg -6 T1:3.2.1-3.2.2 T2:3.4, R2:5.2-6 T1: 3.3.1-5, R5:pg-13 R2:3.1-3 T1: 3.4.1-3.4.3 R2:4.1-4,R3:2.5.3 R4:pg-57 T1: 4.2 T1: 3.5 T1: 4.1

Data warehouse Architecture

14

Data warehouse implementation Further Development of Data cube Technology From data warehousing to data mining Efficient methods for data cube computation

15

16 17 18

Further Development of Data cube Technology Attribute oriented induction

19

Discovery Driven Exploration of data cubes, Complex Aggregation at multiple granularities ,Other development

T1: 4.2.1-3,

20

21 Mining frequent patterns , association and correlation: Basic concepts Efficient and scalable frequent itemset mining methods 22

Attribute oriented induction, Efficient T1: 4.3.1-5, implementation of attribute oriented induction Presentation of the derived generalization ,Mining class comparison, Class description Revision of Unit-II UNIT III Market basket analysis ,Frequent item sets, closed item sets and Association Rule Frequent pattern mining The Apriori algorithm , Generating association rules from frequent item sets Improving the efficiency of apriori, Mining frequent itemsets without candidate generation Mining frequent itemsets using vertical data format Mining closed frequent itemsets Multilevel association rules, Mining multidimensional Association rules from relational databases and data warehouses Strong rules are not necessarily interesting From association analysis to correlation analysis Meta rule Guided mining of association rules Mining guided by additional rule constraints UNIT IV Preparing the data for classification and prediction Comparing classification methods Decision tree induction, Attribute selection measures Tree pruning , Scalability and decision tree induction T1: 5.1.1,2,3 T2:6.1, R3:7.1 T1: 5.2.1-5 T2:6.2-3 R1:4.4,8 R3:7.3 T1:5.2.6 T2:6.5 T1: 5.3.1,2

23

24 Mining various kinds of association rules From association mining to correlation analysis Constraint based association mining Classification and prediction: Issues regarding classification and prediction Classification by Decision tree introduction 25

26 27

T1: 5.4.1,2 T1: 5.5.1,2 R1:4.13 T1: 6.2.1,2 T2:4.1,2, R1:6.1,2

28

29

Bayesian Classification

30

T1: 6.3.1-4 T2:4.3 R1:6.3-6.7 R3:4.7 Bayes theorem, Naive Bayesian classification T1: 6.4.1-4 Bayesian Belief Networks, Training Bayesian Belief T2:5.3 networks R3:9.2 Using If then rules for classification, Rule extraction from a decision tree, Rule induction using a sequential covering algorithm A Multilayer feed-forward neural network, Defining a network topology Back propagation, Back propagation, Back propagation and interpretability Linearly separable, Linearly inseparable Classification by association rule analysis T1: 6.5.1-3 T2:5.1 T1: 6.6.1-4

Rule-based classification

31

Classification of Back propagation Support vector machines Associative Classification

32

33

T1: 6.7.1,2,6.8 T2:5.5 R3:10.1 T1: 6.9.1,2, 6.10.1,2,3 T2:5.2 T1: 6.11.1,2,3

Lazy learners Other classification methods Prediction

34

35

K-nearest neighbor classifiers ,Case-based reasoning Genetic algorithms, Rough set approach, Fuzzy set approaches Linear and multiple regression, Non-liner regression

Accuracy and error measures Evaluating the accuracy of classifier or predictor Ensemble methods- increasing the accuracy

36 37

38

Other regression models Classifier accuracy measures, Predictor error T1: 6.12.1,2 measures Holdout method and random subsampling, Cross- T1: 6.13.1,2,3 validation, Bootstrap T2:4.5 R3:4.8.1 Bagging , Boosting T1: 6.14.1,2 T2:5.6.4,5 Revision of Unit-III & IV UNIT V Interval scaled variables, Binary variables Nominal, ordinal, and ratio scaled variables Variables of mixed types, Vector objects

39 Cluster analysis introduction: Types of Data in cluster analysis, A Categorization of major clustering methods Partitioning methods Hierarchical methods 40

T1: 7.2,3 T2:8.1.1,2 R1:5.1,2

41

Classical partitioning methods, Partitioning methods T1: 7.4.1,2, 7.5.1-4 in large databases T2:8.2, R1:5.3 Agglomerative and divisive , BIRCH, ROCK, R1: 5.4,7,9,13 Chameleon R3:11.2, 11.5 DBSCAN: A Density-based clustering method , T1: 7.6.1,2,3 OPTICS , DENCLUE T2:8.4,9.3.3 R1:5.8, R3:11.6,7 STING , Wave cluster T1: 7.7.1-7.7.2 Expectation maximization, Conceptual clustering T1: 7.8.1,2,3 Neural network approach CLIQUE, PROCLUS, Frequent pattern based T1: 7.9.1,2,3 clustering methods Clustering with obstacle objects, User constraint T1: 7.10.1,2,3 cluster analysis, Semi supervised cluster analysis Statistical distribution based, Distance based, Density T1: 7.11.1-4 based, Deviation based Revision of Unit-V UNIT VI Methodologies for stream data processing and stream T1: 8.1.1-8.1.5, data systems, Stream OLAP and stream data cubes Frequent patterns mining in data streams, Classification of dynamic data streams, Clustering evolving data streams Trend analysis , Similarity search in time-series T1: 8.2.1,2 analysis T2:9.11 sequential pattern mining, scalable methods for T1: 8.3.1-4, mining, constraint based periodicity analysis for time related data alignment of biological sequences, hidden Markov T1: 8.4.1,2 model methods for mining frequent subgraphs , mining T1: 9.1.1,2,3 variant and constrained substructure patterns applications

Density Based methods Grid based methods and Model based clustering methods Clustering high dimensional data Constraint based cluster analysis Outlier analysis

42

43

44 45 46 47

Mining streams, time series and sequence data: Mining data streams

48

Mining time series and sequence data Mining sequence patterns in transactional databases Mining sequence patterns in biological data Graph mining

49 50

51 52

Social network analysis Multirelational data mining

53 54

what is social networks characteristics, link mining mining on social networks what is multirelational datamining? ILP approach, Tuple ID propagation , Multirelational classification Multirelational clustering Revision of Unit-VI

T1: 9.2.1-4, T1: 9.3.1-5,

55 Mining objects,spatial multimedia , text, and web data: Multidimensional analysis and descriptive mining of complex data objects 56

UNIT VII Generalization of structured data , Aggregation and T1: 10.1.1-3 approximation in spatial and multimedia data generalization , Generalization of object identifiers and class Generalization of class composition hierarchies Construction and mining of object cubes Generalization based mining of plan databases by divide and conquer Spatial data cube construction and spatial OLAP Spatial association analysis, clustering methods, classification and trend analysis, Mining Raster databases Similarity search in multimedia data, Multidimensional analysis , Classification and prediction analysis , Mining association in multimedia data Text data analysis and information retrieval Dimensionality reduction, Text mining Mining the webs link structures to identify Automatic classification of web documents Construction of a multilayered web information base, Web usage mining UNIT VIII For financial data analysis, For retail industry For telecommunication industry, For biological data analysis, Other scientific applications, For intrusion detection How to choose a data mining system Examples of commercial data mining systems Revision of Unit-VII & VIII Subject Summary T1: 10.1.3-6

57

Spatial data mining

58

T1:10.2.1 5, R1:9.12-14 R3:12.1 T1: 10.3.1 5, Pg: 607-613

Multimedia data mining

59

Text Data mining Mining the world wide web

60 61

T1: 10.4.1-3 R1:8.6 T1: 10.5.1-5 R1:8.2-8.5

Application and trends in data mining: Data mining application

62 63

T1: 11.1.1,2 R1:3.11 T1: 11.1.3,4, 11.1.5,6 T1: 11.2.1 T1: 11.2.2

Data mining system products and research prototypes

64 65 66 67

Books Referred by faculty: (i) TEXT BOOKS: T1. Jiawei Han & Micheline Kamber. Data Mining Concepts and Techniques, Kaufman Publishers

T2. Tan, Pang-Ning and other Introduction to Data Mining Pearson Education, 2006.
(ii) REFERENCES:

1. Arun K Pujari, Data Mining Techniques, University Press. 2. Sam Anahory & Dennis Murray, Data Warehousing in the Real World, Pearson Edn Asia. 3. K.P.Soman, Insight Into Data Mining, PHI 2008. 4. Paulraj Ponnaiah, Data Warehousing Fundamentals, Wiley Student Edition. 5. Ralph Kimball, The Data Warehouse Life cycle Tool kit, Wiley Student Edition.

WEBSITES:
1. www.the-data-mine.com/bin/view/Misc/IntroductionToData Mining : Links about different journals of Data Warehousing and Mining, Links to Books for Data Mining 2. http://Data Mining warehousing.blogspot.com 3. www.pcc.qub.ac.uk/tec/courses/Data Mining /ohp/dm-OHP-final_1.html: Online notes for DM 4. http://pagesperso-orange.fr/bernard.lupin/english/: Provides information about OLAP 5. http://www.cs.sfu.ca/~han/dmbook : Slides of text book

JOURNALS: 1. IEEE Transactions on Knowledge and Data Engineering: This Journal has papers published on the data mining techniques like classification, Association Analysis, Prediction, Cluster analysis which gives an exposure to students and also they come to know that how these techniques can be used in the real world applications. IEEE Transactions on Pattern Analysis and Machine Intelligence: This includes all traditional areas of computer vision and image understanding, all traditional areas of pattern analysis and recognition, and selected areas of machine intelligence. Areas of such machine learning, search techniques, document and handwriting analysis, medical image analysis, video and image sequence analysis, content-based retrieval of image and video, face and gesture recognition and relevant specialized hardware and/or software architectures are also covered. Data Mining and Knowledge Discovery: The journal publishes original technical papers in both the research and practice of data mining and knowledge discovery, surveys and tutorials of important areas and techniques, and detailed descriptions of significant applications. Its Coverage includes: Theory and Foundational Issues, Data Mining Methods, Algorithms for Data Mining, Knowledge Discovery Process, and Application Issues. Journal of Intelligent Systems : This journal provides readers with a compilation of stimulating and up-to-date articles within the field of intelligent systems. The focus of the journal is on high quality research that addresses paradigms, development, applications and implications in the field of intelligent systems. The list of topics spans all the areas of modern intelligent systems such as: Artificial Intelligence, Pattern Recognition, , Data Mining, Evolutionary algorithms, Swarm Intelligence, Intelligent Agents and Multi-Agent Systems, Decision Support Systems, Supervised Semi-Supervised and Unsupervised Learning, , Knowledge Management and Representation, Intelligent System Design, Bayesian Learning, , Evolving Clustering Methods, Natural Language Processing and Fusion.

2.

3.

4.

STUDENT SEMINAR TOPICS:

1. Integrating Data Warehouses With Web Data, J.M. Perez, R. Berlanga, M.J. Aramburu &T.B.Pedersen, July 2008, Vol.No.20, Pg.No.940. 2. IDD: A Supervised Interval Distance - Based Method for Discretization, F.J. Ruiz, C. Angulo & N. Agell, September 2008, Vol. No. 20, Pg. No. 1230. 3. Automatic Website Summarization By Image Content, J.Li.R.C.-W.Wong, A.W. FU & J.Pei, Sep 2008, Vol. 20, Pg.No.1195. 4. Label Propagation through Linear Neighborhoods Fei Wang, Changshui Zhang, January 2008, Vol.No.20, Pg.No.55. 5. Time-Aware Web Users Clustering, S.G. Petridou, V.A. Koutsonikola, A.I. Vakali & G.I. Papadimitriou, May 2008, Vol.No.20, Pg.No.653. 6. Efficient Similarity Search Over Future Stream Time Series, Xiang Lian, Lei Chen; January 2008, Vol.No.20, Pg.No.40. 7. Distributed Decision Tree Induction within the Grid Data Mining Framework GridMiner -Core, Jurgen Hofer and Peter Brezany 8. Distributed Data Mining in Credit Card Fraud Detection Philip K Chan

QUESTION BANK: UNIT-I 1. (a) Briefly discuss about data integration. (b) Briefly discuss about data transformation. (R07 APRIL 2011) 2. (a) Explain about the graph displays of basic statistical class description. (b) Briefly explain about the presentation of class comparison descriptions. (R07 APRIL 2011) 3. Discuss about primitives for specifying a data mining task. (R07 APRIL 2011) 4. (a) Justify the role of data cube aggregation in data reduction process with an example. (b) Discuss the role of Numerosity reduction in data reduction process in detail. (R07 APRIL 2011) 5. Explain the syntax for the following data mining primitives: (a) Task-relevant data (b) The kind of knowledge to be mined (c) Background knowledge (d) Interestingness measures. (R07 APRIL 2011) 6. Explain the following terms in detail. (a) Concept description (b) Variance and Standard deviation. (c) Mean, median, and mode. (d) Quartiles, outliers, and boxplots. (R07 APRIL 2011) 7. (a) Briefly explain about the forms of Data preprocessing. (b) Discuss issues to be considered during data integration process. (R07 APRIL 2011) 8. (a) Briefly discuss about specifying the kind of knowledge to be mined. (b) Explain the syntax for specifying the kind of knowledge to be mined. (R07 APRIL 2011) 9. (a) Explain data mining as a step in the process of knowledge discovery. (b) Differentiate operational database systems and data warehousing. (R07 APRIL 2011) 10. Discuss about the role of data integration and transformation in data preprocessing. (R07 APRIL 2011) 11. (a) Describe why is it important to have a data mining query language. (b) Briefly discuss about the architectures of data mining systems. (R07 APRIL 2011) 12.a) Explain the storage models of OLAP? b) How does the data warehousing and data mining work together. (R07 APRIL/MAY 2011) 13. Suppose that the data for analysis includes the attribute age. the age values for the data tuples are increasing order 13 16 16 23 23 25 25 25 25 30 30 30 30 35 35 35 40 40 45 45 45 70 a) How might you determine the outliers in the data? b) What other methods are there for data smoothing? (R07 APRIL/MAY 2011) 14. List and describe the primitives for the data mining task? (R07 APRIL/MAY 2011) 15. Write short notes on: i) Discriminating different classes ii) Statistical measures in large databases. (R07 APRIL/MAY 2011) 16.a) Discuss various issues in data integration? b) Explain the concept hierarchy generation for categorical data? (R07 APRIL/MAY 2011) 17.a) Why is it important to have a data mining query language? b) Define schema and operation-derived hierarchies? (R07 APRIL/MAY 2011) 18. List and describe the various types of concept hierarchies? (R07 APRIL/MAY 2011) 19. List the statistical measures for the characterization of data dispersion, and discuss how they can be computed efficiently in large data bases? (R07 APRIL/MAY 2011)

20. What are the various issues in data mining? Explain each one in detail? (R07 APRIL/MAY 2011) 21. Why preprocess the data and explain in brief? (R07 APRIL/MAY 2011) 22. Write short notes on GUI, DMQL? How to design GUI based on DMQL? (R07 APRIL/MAY 2011) 23.a) What are the various issues relating to the diversity of database types? b) Explain how data mining used in health care analysis? (R07 APRIL/MAY 2011) 24. (a) When is a summary table too big to be useful ? (b) Relate and discuss the various degrees of aggregation within summary tables. (NR APRIL 2011) 25. (a) Explain the design and construction process of data warehouses. (b) Explain the architecture of a typical data mining system. (JNTU Dec-2010) 26. (a) How can we smooth out noise in data cleaning process? Explain. (b) Why preprocessing of data is needed? (JNTU Dec-2010) 27. (a) Briefly discuss the data smoothing techniques. (b) Explain about concept hierarchy generation for categorical data. (JNTU Dec-2010) 28. (a) Explain data mining as a step in the process of knowledge discovery. (b) Differentiate operational database systems and data warehousing. (JNTU Dec-2010) 29. (a) How can you go about filling in the missing val ues in data cleaning process? (b) Discuss the data smoothing techniques. (JNTU Dec-2010) 30. Write short notes on the following: (a) Association analysis (b) Classiffication and prediction (c) Cluster analysis (d) Outlier analysis. (JNTU Dec-2010) 31. (a) What are the desired architectures for Data mining systems. (b) Briey explain about concept hierarchies. (JNTU Dec 2010) 32. (a) List and describe any four primitives for specifying a data mining task. (b) Describe why concept hierarchies are useful in data mining. (JNTU Dec-2010) 33. Briey discuss the following data mining primitives: (a) Task-relevant data (b) The kind of knowledge to be mined (c) Interestingness measures (d) Presentation and visualization of discovered patterns. (JNTU Dec-2010) 34. (a) Briefly discuss about data integration. (b) Briefly discuss about data transformation. (JNTU Dec-2010) 35. (a) Discuss construction and mining of object cubes. (b) Give a detail note on trend analysis. (JNTU Dec-2010) 36. (a) Explain the design and construction process of data warehouses. (b) Explain the architecture of a typical data mining system. (JNTU Dec-2010) 37. (a) Draw and explain the architecture of typical data mining system. (b) Differentiate OLTP and OLAP. (JNTU Aug/Sep 2008) 38. (a) Explain data mining as a step in the process of knowledge discovery. (b) Differentiate operational database systems and data warehousing.(JNTU Aug/Sep 2008) 39. (a) Explain the architecture of a typical data mining system. (b) Discuss the issues regarding data warehouse architecture. (JNTU Aug/Sep 2008) 40. (a) Briefly discuss the data smoothing techniques. (b) Explain about concept hierarchy generation for categorical data. (JNTU Aug/Sep 2008) 41. Explain various data reduction techniques. (JNTU Aug/Sep 2008) 42. (a) List and describe any four primitives for specifying a data mining task. (b) Describe why concept hierarchies are useful in data mining. (JNTU Aug/Sep 2008)

43. (a) Briefly discuss various forms of presenting and visualizing the discovered patterns. (b) Discuss about the objective measures of pattern interestingness. (JNTU Aug/Sep 2008) 44. Write the syntax for the following data mining primitives. (a) The kind of knowledge to be mined. (b) Measures of pattern interestingness. (JNTU Aug/Sep 2008, May 2008) 45. (a) How can we specify a data mining query for characterization with DMQL? (b) Describe the transformation of a data mining query to a relational query. (JNTU Aug/Sep 2008) 46. Briefly compare the following concepts. Use an example to explain your points. (a) Snowflake schema, fact constellation, starnet query model. (b) Data cleaning, data transformation, refresh. (c) Discovery driven cube, multifeature cube, and virtual warehouse. (JNTU Aug/Sep 2008) 47. Write the syntax for the following data mining primitives: (a) Task-relevant data. (b) Concept hierarchies. (JNTU May 2008) 48. (a) Describe why is it important to have a data mining query language. (b) The four major types of concept hierarchies are: schema hierarchies, set- grouping hierarchies, operation-derived hierarchies, and rule-based hierarchies. Briefly define each type of hierarchy. (JNTU May 2008) 49. (a) Explain the syntax for Task-relevant data specification. (b) Explain the syntax for specifying the kind of knowledge to be mined.(JNTU May 2008) 50. (a) Draw and explain the architecture for on-line analytical mining. (b) Briefly discuss the data warehouse applications. (JNTU May 2008) 51. (a) Explain data mining as a step in the process of knowledge discovery. (b) Differentiate operational database systems and data warehousing. (JNTU May 2008) 52. (a) Explain the major issues in data mining. (b) Explain the three-tier datawarehousing architecture. (JNTU May 2008) 53. (a) Explain data mining as a step in the process of knowledge discovery. (b) Differentiate operational database systems and data warehousing. (JNTU May 2008) 54. Briefly discuss the role of data cube aggregation and dimension reduction in the data reduction process. (JNTU May 2008) 55. (a) Briefly discuss about data integration. (b) Briefly discuss about data transformation. (JNTU May 2008) 56. Discuss the role of data compression and numerosity reduction in data reduction process. (JNTU May 2008) UNIT-II 1. (a) What is Concept description? Explain. (b) What are the differences between concept description in large data bases and OLAP? (R07 APRIL 2011) 2. (a) What does the data warehouse provide for business analyst? Explain (b) How do data warehousing and OLAP related to Data mining? (R07 APRIL 2011) 3. (a) Draw and explain the architecture of typical data mining system. (b) Differentiate OLTP and OLAP. (R07 APRIL 2011) 4. Write short notes for the following in detail: (a) Attribute-oriented induction. (b) Efficient implementation of Attribute-oriented induction. (R07 APRIL 2011) 5. Briefly compare and explain by taking an example of your point(s). a) Snowflake schema, fact constellation b) Data cleaning, data transformation. (R07 APRIL/MAY 2011) 6.a) Differentiate between OLAP and OLTP?

b) Draw and explain the star schema for the data warehouse? (R07 APRIL/MAY 2011) 7. What is data compression? How would you compress data using principle component analysis (PCA)? (R07 APRIL/MAY 2011) 8. How is class comparison performed? Can class comparison mining be implemented efficiently using data cube techniques? If yes explain? (R07 APRIL/MAY 2011) 9. (a) Explain the ADHOC query and Automation in Data Warehouse delivery pro-cess. (b) Explain to the idea Can we do without an Enterprise data warehouse"? (NR APRIL 2011) 10. (a) Explain the basic levels of testing a data warehouse. (b) Explain a plan for testing the data warehouse. (NR APRIL 2011) 11. (a) Describe the server management features of a data warehouse system. (b) Management tools are required to manage a large, dynamic and complex system such as data warehouse system" ower your explanation with justification. (NR APRIL 2011) 12. Explain the significance of tuning the Data warehouse. (NR APRIL 2011) 13. (a) Discuss the design strategies to implement backup strategies. (b) Describe the recovery strategies of a data warehouse system. (NR APRIL 2011) 14. (a) How can we perform discrimination between different classes? Explain. (b) Explain the analytical characterization with an example. (JNTU Dec 2010) 15. Write short notes for the following in detail: (a) Measuring the central tendency (b) Measuring the dispersion of data. (JNTU Dec-2010) 16. Suppose that the following table is derived by Attribute-oriented induction. CLASS BIRTH-PLACE COUNT Canada 180 Programmer others 120 Canada 20 DBA others 80 (a) Transform the table into a crosstab showing the associated t-weights and d-weights. (b) Map the class programmer into a(Bi-directional) quantitative descriptive rule,for example, 8X,Programmer(X) , (birth place(X)= "Canada" ^ ...) [t: x%,d:y%] ... ^ (...) [ t:w%, d:z%] (JNTU Dec-2010) 17. (a) Differentiate attribute generalization threshold control and generalized relation threshold control. (b) Differentiate between predictive and descriptive data mining. (JNTU Dec-2010) 18. (a) What are the differences between concept description in large data bases and OLAP? (b) Explain about the graph displays of basic statistical class description. (JNTU Aug/Sep 2008) 19. (a) How can we perform attribute relevant analysis for concept description? Explain. (b) Explain the measures of central tendency in detail. (JNTU Aug/Sep 2008, May 2008) 20. Write short notes for the following in detail: (a) Measuring the central tendency (b) Measuring the dispersion of data. (JNTU May 2008) 21. (a) Write the algorithm for attribute-oriented induction. Explain the steps involved in it. (b) How can concept description mining be performed incrementally and in a distributed manner? (JNTU May 2008) 22. (a) What are the differences between concept description in large data bases and OLAP? (b) Explain about the graph displays of basic statistical class description. (JNTU May 2008)
UNIT-III

1. Give a detail note on classiffication based on concepts from association rule mining. APRIL 2011) 2. (a) Explain how concept hierarchies are used in mining multilevel association rule?

(R07

(b) Give the classiffication of association rules in detail. (R07 APRIL 2011) 3. (a) How sclable is decision tree induction?Eplain (b) Discuss classiffication based on concept from association rule mining. (R07 APRIL 2011) 4.a) How is association rules mined from large databases? b) Describe the different classifications of associate rule mining? (R07 APRIL/MAY 2011) 5. List and explain the five techniques to improve the efficiency apriori algorithm? (R07 APRIL/MAY 2011) 6. What is Divide and Conquer? How it could be helpful for FP Growth method in generating frequent item sets without candidate generation? (R07 APRIL/MAY 2011) 7. Describe example of data set for which apriori check would actually increase the cost? (R07 APRIL/MAY 2011)
8.Compare and contrast Apriori algorithm with frequent pattern growth algorithm. Consider a data set apply both algorithms and explain the results. (JNTU Dec 2010) 9.a) Discribe mining multidimensional association rule using static discretization of quantitative attribute. b) Expalin association rule generation from frequent itemsets. (JNTU Dec-2010) 10..Explain mining multilevel association rules from transaction databases. (JNTU Dec-2010) 11.(a) Explain how concept hierarchies are used in mining multilevel association rule? (b) Give the classification of association rules in detail. (JNTU Dec-2010) 12.Sequential patterns can be mined in methods similar to the mining of association rules. Design an efficient algorithm to mine multilevel sequential patterns from a transaction database. An example of such a pattern is the following A customer who buys a PC will buy Microsoft software within three moths on which one may drill down to find a more refined version of the patters, such as A customer who buys a Pentium PC will buy Microsoft office within three months. (JNTU Aug/Sep 2008) 13. (a)Which algorithms an influential algorithm for mining frequent item sets for Boolean association rules? Explain. (b)What are additional rule constraints to guide mining? Explain. (JNTU Aug/Sep 2008) 14. (a) Discuss about Association rule mining. (b) What are the approaches for mining multilevel Association rules? Explain. (JNTU Aug/Sep 2008) 15. Explain the Apriori algorithm with example. (JNTU Aug/Sep 2008, May 2008) 16. (a) Write the FP-growth algorithm. Explain. (b) What is an iceberg query? Explain with example. (JNTU May 2008) 17. (a) How can we mine multilevel Association rules efficiently using concept hierarchies? Explain. (b) Can we design a method that mines the complete set of frequent item sets without candidate generation. If yes, explain with example. (JNTU May 2008) UNIT-IV

1. (a) Discuss about Concept hierarchy. (b) Briefly explain about - classiffication of database systems. (R07 APRIL 2011) 2. (a) Explain Distance-based discretization. (b) Give a detail note on iceberg queries . (R07 APRIL 2011) 3. (a) Discuss the various measures available to judge a classiffier. (b) Give a note on neive Bayesian classiffier. (R07 APRIL 2011) 4. (a) Discuss automatic classiffication of web documents . (b) Write about basic measures of text retrieval. (c) Explain mining raster databases. [5+6+5] (R07 APRIL 2011) 5. Explain in detail the major steps of decision tree classiffication. (R07 APRIL 2011) 6. Why perform attribute relevance analysis? Explain the various methods of its? (R07 APRIL/MAY 2011) 7. How will you solve a classification problem using decision trees? (R07 APRIL/MAY 2011)

8. Outline a data cube-based incremental algorithm for mining analytical class comparisons? (R07 APRIL/MAY 2011) 9. What is backpropagation? Explain classification by back-propagation? (R07 APRIL/MAY 2011) 10. Can we get classification rules from decision trees? If so how? What are the enhancements to the basic decision tree? (R07 APRIL/MAY 2011) 11. Explain the various preprocessing steps to improve the accuracy, efficiency, and scalability of the classification or prediction process? (R07 APRIL/MAY 2011) 12.Explain in detail the major steps of decision tree classiffication. (JNTU Dec 2010) 8. (a) Discuss the five criteria for the evaluation of classification and prediction methods. (b) Explain how rules can be extracted from training neural networks. (JNTU Dec-2010) 9. (a) Explain with an example a measure of the goodness of split. (b) Write a detail note on genetic algorithms for classiffication. (JNTU Dec-2010) 15. (a) Give a note on log-linear models. (b) Explain the hold out method for estimating classiffier accuracy. (c) Discuss Fuzzy set approach for classiffication. [5+5+6] (JNTU Dec-2010) 16. Discuss about Back propagation classification. (JNTU Aug/Sep 2008) 17. The following table consists of training data from an employee database. The data have been generalized. For a given row entry, count represents the number of data tuples having the values for department, status, age and salary given in that below.

18. 19.

20.

21.

22.

Let salary be the class label attribute. Design a multiplayer feed-forward neural network for the given data. Label the nodes in the input and output layers. (JNTU Aug/Sep 2008) (a) Explain decision tree induction classification. (b) Describe backpropagation classification. (JNTU Aug/Sep 2008. May 2008) (a) Can any ideas from association rule mining be applied to classification? Explain. (b) Explain training Bayesian belief networks. (c) How does tree pruning work? What are some enhancements to basic decision tree induction? (JNTU Aug/Sep 2008) (a) What is classification? What is prediction? (b) What is Bayes theorem? Explain about Naive Bayesian classification. (c) Discuss about k-Nearest neighbor classifiers and case-based reasoning. (JNTU May 2008) (a) Describe the data classification process with a neat diagram. (b) How does the Naive Bayesian classification works? Explain. (c) Explain classifier accuracy. (JNTU May 2008) (a) Explain about basic decision tree induction algorithm.

(b) Discuss about Bayesian classification. UNIT-V

(JNTU May 2008)

1. (a) Explain competitive learning and self organizing feature maps methods to clustering. (b) Discuss in detail BIRCH algorithm. (R07 APRIL 2011) (b) What is an outlier? Explain in brief outlier analysis. (R07 APRIL 2011) each. (R07 APRIL 2011) 4. What is association analysis? Discuss cluster analysis. Explain the correlation between these two types of analysis. (R07 APRIL 2011) 5. (a) How are association rules mined from large databases? Explain. (b) Explain in detail constraint based association mining. (R07 APRIL 2011) 6. (a) Give a detail note on CLIQUE algorithm. (b) Discuss expectation maximization algorithm for clustering. (R07 APRIL 2011) 7.a) What are the fields in which clustering techniques are used? b) What are the major requirements of clustering analysis? (R07 APRIL/MAY 2011) 8. Why is outlier mining important? Discuss about different outlier detection approaches? Briefly discuss about any two hierarchical clustering methods with suitable examples? (R07 APRIL/MAY 2011) 9. What are the different types of data used in cluster analysis? Explain in brief each one with an example? (R07 APRIL/MAY 2011) 10.a) What are the differences between clustering and nearest neighbor prediction? b) Define nominal, ordinal, and ratio scaled variables? (R07 APRIL/MAY 2011)
11. (a) Explain how COBWEB method is used for clustering. (JNTU Dec 2010) (b) Discuss in detail DENCLUE clustering methods. 12. (a) Explain K-means algorithm for clustering. (b) Given two objects represented by the tuples(22,1,42,10) and(20,0,36,8) i. Compute the Manhatten distance between the two objects. ii. Compute the Euchidean distance between the two objects. (JNTU Dec-2010) 13. (a) Discuss interval-scaled vari abl e and binary vari abl es. (b) Explain in detail K-Medoids algorithm for clustering. (JNTU Dec-2010) 14. (a) Discuss distance based outlier detection. (b) Explain OPTICS algorithm for clustering. (JNTU Dec-2010) 15. (a) What major advantages does DENCLUE have in comparison with other clustering algorithms? (b) What advantages does STING offer over other clustering methods? (c) Why wavelet transformation useful for clustering? (d) Explain about outlier analysis. (JNTU Aug/Sep 2008) 16. (a) Define mean absolute deviation, z-source, city block distance, and minikowski distance. (b) What are different types of hierarchical methods? Explain. (JNTU Aug/Sep 2008) 17. (a) Discuss about binary, nominal, ordinal and ratio-scaled variables. (b) Explain about grid-based methods. (JNTU Aug/Sep 2008) 18. (a) What are the categories of major clustering methods? Explain. (b) Explain about outlier analysis. (JNTU Aug/Sep 2008) 19. (a) Given the following measurement for the variable age: 18, 22, 25, 42, 28, 43, 33, 35, 56, 28 Standardize the variable by the following: i. Compute the mean absolute deviation of age. ii. Compute the Z-score for the first four measurements.

(b) What is a distance-based outlier? What are efficient algorithms for mining distance-based algorithm? How are outliers determined in this method? (JNTU May 2008) 20. (a) Write algorithms for k-Means and k-Medoids. Explain. (b) Discuss about density-based methods. (JNTU May 2008) 21. (a) Given two objects represented by the tuples (22,1,42,10) and (20,0,36,8): i. Compute the Euclidean distance between the two objects. ii. Compute the Manhanttan distance between the two objects. iii. Compute the Minkowski distance between the two objects, using q=3. (b) Explain about Statistical-based outlier detection and Deviation-based outlier detection. (JNTU May 2008) UNIT-VI

1. (a) Describe the essential features of temporal data and temporal inferences. (b) Discuss the major algorithms of the sequence mining problem. (NR APRIL 2011)
UNIT-VII

1. (a) Discuss various ways to estimate the trend. (b) Explain construction of a multilayered web information base. (R07 APRIL 2011) (b) Explain latent semantic inducing technique. (R07 APRIL 2011) 3. (a) Discuss data transformation from time domain to frequency domain. (b) Explain HITS algorithm for web structure mining. (R07 APRIL 2011) 4. Write short notes on: i) Mining Spatial Databases ii) Mining the World Wide Web. (R07 APRIL/MAY 2011) 5. Write short notes on: i) Data objects ii) Sequence Data Mining iii) Mining Text Databases. (R07 APRIL/MAY 2011) 6. (a) Which frequent itemset mining is suitable for text mining and why. Explain? (b) Discuss the relationship between text mining and information retrieval and information extraction. (NR APRIL 2011) 7. (a) Describe cosine measure for similarity in documents. (JNTU Dec 2010)
(b) Explain in detail similarity search in time-series analysis. 8. (a) Explain spatial datacube construction and spatial OLAP. (b) Give a note on item frequency matrix. (JNTU Dec 2010) 9. (a) Discuss web content mining and web usage mining. (b) compare information retrieval with text mining. (JNTU Dec 2010) 10. (a) Explain spatial data cube construction and spatial OLAP. (b) Discuss about mining text databases. (JNTU Aug/Sep 2008) 11. (a) Define spatial database, multimedia database, time-series database, sequence database, and text database. (b) What is web usage mining? Explain with suitable example. (JNTU Aug/Sep 2008) 12. A heterogeneous database system consists of multiple database systems that are defined independently, but that need to exchange transform information among themselves and answer global queries. Discuss how to process a descriptive mining query in such a system using a generalizationbased approach. (JNTU Aug/Sep 2008) 13. (a) How to mine Multimedia databases? Explain. (b) Define web mining. What are the observations made in mining the web for effective resource and knowledge discovery? (c) What is web usage mining? (JNTU Aug/Sep 2008)

.14. An e-mail database is a database that stores a large number of electronic mail messages. It can be viewed as a semistructured database consisting mainly of text data. Discuss the following. (a) How can such an e-mail database be structured so as to facilitate multi-dimensional search, such as by sender, by receiver, by subject, by time, and so on? (b) What can be mined from such an e-mail database? (c) Suppose you have roughly classified a set of your previous e-mail messages as junk, unimportant, normal, or important. Describe how a data mining system may take this as the training set to automatically classify new e-mail messages for unclassified ones (JNTU May 2008) 15. Suppose that a city transportation department would like to perform data analysis on highway traffic for the planning of highway construction based on the city traffic data collected at different hours every day. (a) Design a spatial data warehouse that stores the highway traffic information so that people can easily see the average and peak time traffic flow by highway, by time of day, and by weekdays, and the traffic situation when a major accident occurs. (b) What information can we mine from such a spatial data warehouse to help city planners? (c) This data warehouse contains both spatial and temporal data. Propose one mining technique that can efficiently mine interesting patterns from such a spatio-temporal data warehouse (JNTU May 2008) 16. Explain the following: (a) Constriction and mining of object cubes (b) Mining associations in multimedia data (c) Periodicity analysis (d) Latent semantic indexing (JNTU May 2008) 17. (a) Give an example of generalization-based mining of plan databases by divide- and-conquer. (b) What is sequential pattern mining? Explain. (c) Explain the construction of a multilayered web information base. (JNTU May 2008)

UNIT-VIII

ASSIGNMENT QUESTIONS: UNIT-I Unit-I 1. Describe three challenges to data mining regarding data mining methodology and user interaction issues. 2. What are the major challenges of mining a huge amount of data (such as billions of tuples) in comparison with mining a small amount of data (such as a few hundred tuple data set)? 3. Outline the major research challenges of data mining in one specific application domain, such as stream/sensor data analysis, spatiotemporal data analysis, or bioinformatics 4. What is the difference between discrimination and classification? Between characterization and clustering? Between classification and prediction? For each of these pairs of tasks, how are they similar? 5. Outliers are often discarded as noise. However, one persons garbage could be anothers treasure. For example, exceptions in credit card transactions can help us detect the fraudulent use of credit cards. Taking fraudulence detection as an example, propose two methods that can be used to detect outliers and discuss which one is more reliable. 6. Recent applications pay special attention to spatiotemporal data streams. A spatiotemporal data stream contains spatial information that changes over time, and is in the form of stream data (i.e., the data flow in and out like possibly infinite streams). (a) Present three application examples of spatiotemporal data streams. (b) Discuss what kind of interesting knowledge can be mined from such data streams, with limited time and resources. (c) Identify and discuss the major challenges in spatiotemporal data mining. (d) Using one application example, sketch a method to mine one kind of knowledge from such stream data efficiently. 7. Describe the differences between the following approaches for the integration of a data mining system with a database or data warehouse system: no coupling, loose coupling, 8. Present an example where data mining is crucial to the success of a business .What data mining functions does this business need? Can they be performed alternatively by data query processing or simple statistical analysis? 9. Give three additional commonly used statistical measures (i.e., not illustrated in this chapter) for the characterization of data dispersion, and discuss how they can be computed efficiently in large databases. 10. In many applications, new data sets are incrementally added to the existing large data sets. Thus an important consideration for computing descriptive data summary is whether a measure can be computed efficiently in incremental manner. Use count, standard deviation, and median as examples to show that a distributive or algebraic measure facilitates efficient incremental computation, whereas a holistic measure does not. 11. In real-world data, tuples with missing values for some attributes are a common occurrence. Describe various methods for handling this problem. 12. Use a flowchart to summarize the following procedures for attribute subset selection: (a) stepwise forward selection

(b) stepwise backward elimination (c) a combination of forward selection and backward elimination 13. The median is one of the most important holistic measures in data analysis. Propose several methods for median approximation. Analyze their respective complexity under different parameter settings and decide to what extent the real value can be approximated. Moreover, suggest a heuristic strategy to balance between accuracy and complexity and then apply it to all methods you have given. 14. Explain why it is not possible to analyze some large data sets using classical modeling techniques. Explain the differences between statistical and machine - learning approaches to the analysis of large data sets. 15. Why are preprocessing and dimensionality reduction important phases in successful data - mining applications? 16. Give examples of data where the time component may be recognized explicitly and other data where the time component is given implicitly in a data organization. 17. Why is it important that the data miner understand data well? Give examples of structured, semi - structured, and unstructured data from everyday situations. 18. Can a set with 50,000 samples be called a large data set? Explain your answer. Enumerate the tasks that a data warehouse may solve as a part of the data mining process. 19. Which of the following quantities is likely to show more temporal autocorrelation: daily rainfall or daily temperature? Why? 20. Many sciences rely on observation instead of (or in addition to) designed experiments. Compare the data quality issues involved in observational science with those of experimental science and data mining.

Unit-II 1. Suppose that a data warehouse consists of the four dimensions, date, spectator, location, and game, and the two measures, count and charge, where charge is the fare that a spectator pays when watching a game on a given date. Spectators may be students, adults, or seniors, with each category having its own charge rate. (a) Draw a star schema diagram for the data warehouse. (b) Starting with the base cuboid [date, spectator, location, game], what specific OLAP operations should one perform in order to list the total charge paid by student spectators at GM Place in 2004? (c) Bitmap indexing is useful in data warehousing. Taking this cube as an example, briefly discuss advantages and problems of using a bitmap index structure. 2. A data warehouse can be modeled by either a star schema or a snowflake schema. Briefly describe the similarities and the differences of the two models, and then analyze their advantages and disadvantages with regard to one another. Give your opinion of which might be more empirically useful and state the reasons behind your answer. 3. Design a data warehouse for a regional weather bureau. The weather bureau has about 1,000 probes, which are scattered throughout various land and ocean locations in the region to collect basic weather data, including air pressure, temperature, and precipitation at each hour. All data are sent to the central station, which has collected such data for over 10 years. Your design should facilitate efficient querying and on-line analytical processing, and derive general weather patterns in multidimensional space.

4. A popular data warehouse implementation is to construct a multidimensional database, known as a data cube. Unfortunately, this may often generate a huge, yet very sparse multidimensional matrix. Present an example illustrating such a huge and sparse data cube. 5. State why, for the integration of multiple heterogeneous information sources, many companies in industry prefer the update-driven approach (which constructs and uses data warehouses), rather than the query-driven approach (which applies wrappers and integrators). 6. Describe situations where the query-driven approach is preferable over the update-driven approach. 7. What is a factless fact table? Design a simple STAR schema with a factless fact table to track patients in a hospital by diagnostic procedures and time. 8. You are the data design specialist on the data warehouse project team for a manufacturing company. Design a STAR schema to track the production quantities. Production quantities are normally analyzed along the business dimensions of product, time, parts used, production facility, and production run. State your assumptions. 9. In a STAR schema to track the shipments for a distribution company, the following dimension tables are found: (1) time, (2) customer ship-to, (3) ship-from, (4) product, (5) type of deal, and (6) mode of shipment. Review these dimensions and list the possible attributes for each of the dimension tables. Also, designate a primary key for each table. 10. A dimension table is wide; the fact table is deep. Explain. 11. A data warehouse is subject-oriented. What would be the major critical business subjects for the following companies? (a) an international manufacturing company (b) a local community bank (c) a domestic hotel chain 12. You are the data analyst on the project team building a data warehouse for an insurance company. List the possible data sources from which you will bring the data into your data warehouse. State your assumptions. 13. For an airlines company, identify three operational applications that would feed into the data warehouse. What would be the data load and refresh cycles? 14. Prepare a table showing all the potential users and information delivery methods for a data warehouse supporting a large national grocery chain. 15. Briefly compare the following concepts. You may use an example to explain your point(s). (a) Snowflake schema, fact constellation, star net query model (b) Data cleaning, data transformation, refresh (c) Enterprise warehouse, data mart, virtual warehouse 16. Suppose that a data warehouse consists of the three dimensions time, doctor, and patient, and the two measures count and charge, where charge is the fee that a doctor charges a patient for a visit. (a) Enumerate three classes of schemas that are popularly used for modeling data warehouses. (b) Draw a schema diagram for the above data warehouse using one of the schema classes listed in (a). (c) Starting with the base cuboid [day, doctor, patient], what specific OLAP operations should be performed in order to list the total fee collected by each doctor in 2004?

(d) To obtain the same list, write an SQL query assuming the data are stored in a relational database with the schema fee (day, month, year, doctor, hospital, patient, count, charge). 17. Suppose that a data warehouse for Big University consists of the following four dimensions: student, course, semester, and instructor, and two measures count and avg grade. 18. When at the lowest conceptual level (e.g., for a given student, course, semester, and instructor combination), the avg grade measure stores the actual course grade of the student. At higher conceptual levels, avg grade stores the average grade for the given combination. (a) Draw a snowflake schema diagram for the data warehouse. (b) Starting with the base cuboid [student, course, semester, instructor], what specific OLAP operations (e.g., roll-up from semester to year) should one perform in order to list the average grade of CS courses for each Big University student.

UNIT-III
1. Below database has five transactions. Let min sup = 60% and min con f = 80%.

(a) Find all frequent itemsets using Apriori and FP-growth, respectively. Compare the efficiency of the two mining processes. (b) List all of the strong association rules (with support s and confidence c) matching the following metarule, where X is a variable representing customers, and itemi denotes variables representing items (e.g., A, B, etc.): 2. Consider the following data set. Cust_ID 1 1 2 2 3 3 4 4 5 5 Transaction ID 0001 0024 0012 0031 0015 0022 0029 0040 0033 0038 Items Bought {a, d, e} {a, b, c, e} {a, b, d, e} {a, c, d, e} {b, c, e} {b, d, e} {c, d} {a, b, c} {a, d, e} {a, b, e}

(a) Compute the support for itemsets {e}, {b, d}, and {b, d, e} by treating each transaction ID as a market basket. (b) Use the results in part (a) to compute the confidence for the association rules {b, d} {e} and {e} {b, d}. Is confidence a symmetric measure? (b) Repeat part (a) by treating each customer ID as a market basket. Each item should be treated as a binary variable (1 if an item appears in at least one transaction bought by the customer, and 0 otherwise.) (c) Use the results in part (c) to compute the confidence for the association rules {b, d} {e} and {e} {b, d}. (e) Suppose s1 and c1 are the support and confidence values of an association rule r when treating each transaction ID as a market basket. Also, let s2 and c2 be the support and confidence values of r when treating each customer ID as a market basket. Discuss whether there are any relationships between s1 and s2 or c1 and c2. 3. Implement frequent itemset mining algorithm Apriori (mining using vertical data format), using a programming language that you are familiarwith, such as C++ or Java. Compare the performance of each algorithm with various kinds of large data sets. Write a report to analyze the situations (such as data size, data distribution, minimal support threshold setting, and pattern density) where one algorithm may perform better than the others, and state why. 4. Suppose that a large store has a transaction database that is distributed among four locations. Transactions in each component database have the same format, namely Tj : (i1, ., im), where Tj is a transaction identifier, and ik (1<=k<=m) is the identifier of an item purchased in the transaction. Propose an efficient algorithm to mine global association rules (without considering multilevel associations). You may present your algorithm in the form of an outline. Your algorithm should not require shipping all of the data to one site and should not cause excessive network communication overhead. 5. Implement frequent itemset mining algorithm ECLAT (mining using vertical data format), using a programming language that you are familiar with, such as C++ or Java. Compare the performance of each algorithm with various kinds of large data sets. Write a report to analyze the situations (such as data size, data distribution, minimal support threshold setting, and pattern density) where one algorithm may perform better than the others, and state why. 6. Association rule mining often generates a large number of rules. Discuss effective methods that can be used to reduce the number of rules generated while still preserving most of the interesting rules. 7. What is the essential difference between association rules and decision rules? 8. Given a simple transactional database X:

Using the threshold values support = 25% and confi dence = 60%, (a) find all large itemsets in database X; (b) find strong association rules for database X; (c) analyze misleading associations for the rule set obtained in (b). 9. From the transactional database of question 7, find FP tree for this database if (a) support threshold is 5 (b) support threshold is 3 (c) support threshold is 4 draw the tree step by step. 10. Why is the process of discovering association rules relatively simple compared with generating large itemsets in transactional databases? 11. Solve question 18 with support of 50% and confidence 60%. 12. What are the common values for support and confidence parameters in the Apriori algorithm? Explain using the retail industry as an example. 13. Propose and outline a level-shared mining approach to mining multilevel association rules in which each item is encoded by its level position, and an initial scan of the database collects the count for each item at each concept level, identifying frequent and sub frequent items. Comment on the processing cost of mining multilevel associations with this method in comparison to mining single-level associations. 14. Implement frequent itemset mining algorithm FP-growth (mining using vertical data format), using a programming language that you are familiar with, such as C++ or Java. Compare the performance of each algorithm with various kinds of large data sets. Write a report to analyze the situations (such as data size, data distribution, minimal support threshold setting, and pattern density) where one algorithm may perform better than the others, and state why. 15. Solve question 1 with support of 50% and confidence 60%. 16. Using data in question 2 find all association rules using FP-Growth with a support of 50% and confidence 60%. 17. Suppose we have market basket data consisting of 100 transactions and 20 items. If the support for item a is 25%, the support for item b is 90% and the support for itemset {a, b} is 20%. Let the support and confidence thresholds be 10% and 60%, respectively. (a) Compute the confidence of the association rule {a} {b}. Is the rule interesting according to the confidence measure? (b) Compute the interest measure for the association pattern {a, b}. Describe the nature of the relationship between item a and item b in terms of the interest measure. (c) What conclusions can you draw from the results of parts (a) and (b)?

(d) Prove that if the confidence of the rule {a} {b} is less than the support of {b}, then: i. c({a} {b}) > c({a} {b}), ii. c({a} {b}) > s({b}), where c() denote the rule confidence and s() denote the support of an itemset. 18. Using data in question 2 find all association rules using apriori with a support of 50% and confidence 60%. 19. From the transactional database of question 7, find FP tree for this database if (a) support threshold is 2 (b) support threshold is 3 (c) support threshold is 5 draw the tree step by step. 20. Generate association rules from the following dataset. Minimum support is taken as 22% and minimum confidence is 70%. TID T1 T2 T3 I2, I3 T4 T5 T6 I2, I3 T7 I1, I3 T8 I1, T9 I2 I1, I2, I3

List of I1, I2, I2, I4 Items I5

I1, I2, I1, I3 I4

,I3, I5

UNIT-IV 1) It is difficult to assess classification accuracy when individual data objects may belong to more than one class at a time. In such cases, comment on what criteria you would use to compare different classifiers modeled after the same data. 2) Show that accuracy is a function of sensitivity and specificity, that is, prove this equation

3) Suppose that we would like to select between two prediction models, M1 and M2. We have performed 10 rounds of 10-fold cross-validation on each model, where the same data partitioning in round i is used for both M1 and M2. The error rates obtained for M1 are 30.5, 32.2, 20.7, 20.6, 31.0, 41.0, 27.7, 26.0, 21.5, 26.0. The error rates for M2 are 22.4, 14.5, 22.4, 19.6, 20.7, 20.4, 22.1, 19.4, 16.2, 35.0. Comment on whether one model is significantly better than the other considering a significance level of 1%. 4) Given a decision tree, you have the option of (i) converting the decision tree to rules and then pruning the resulting rules, or (ii) pruning the decision tree and then converting the pruned tree to rules. What advantage does (i) have over (ii)? 5) Take a data set with at least 5 attributes and 15 records and apply decision tree (information gain)classification. 6) Why is nave Bayesian classification called nave? Briefly outline the major ideas of nave Bayesian classification. 7) Take a data set with at least 5 attributes and 15 records and apply decision tree ( gain ratio)classification.

8) The following table shows the midterm and final exam grades obtained for studentsin a database course. x y Midterm exam Final exam 72 84 50 63 81 77 74 78 94 90 86 75 59 49 83 79 65 77 33 52 88 74 81 90 (i) Plot the data. Do x and y seem to have a linear relationship? (ii) Use the method of least squares to find an equation for the prediction of a students final exam grade based on the students midterm grade in the course. (iii) Predict the final exam grade of a student who received an 86 on the midterm exam. 9) Consider the following data set for a binary class problem. A B Class Label T F + T T + T T + T F _ T T + F F _ F F _ F F _ T T _ T F _

(a) Calculate the information gain when splitting on A and B. Which attribute would the decision tree induction algorithm choose? (b) Calculate the gain in the Gini index when splitting on A and B. Which attribute would the decision tree induction algorithm choose? (c)In real-world data, tuples with missing values for some attributes are a common occurrence. Describe various methods for handling this problem.

10) The following table summarizes a data set with three attributes A, B, C and two class labels +, . Build a two-level decision tree.

(a)

According to the classification error rate, which attribute would be chosen as the first splitting attribute? For each attribute, show the contingency table and the gains in classification error rate. (b)Repeat for the two children of the root node. (c)How many instances are misclassified by the resulting decision tree? (d) Repeat parts (a), (b), and (c) using C as the splitting attribute.
11) Using Bayesian method and

Customer ID E1 E2 E3 E4 E5 E6 E7 E8 E9 E10 E11 E12 E13 E14

Height Short Tall Tall Tall Short Tall Average Average Tall Average Tall Short Average Tall

Hair Dark Dark Dark red Blond Blond Blond Blond Grey Grey Blond Blond Grey red

Eyes Blue Black Blue Black Brown Black Blue Blue Blue Black Brown Blue Brown Brown

Credit Rating A B B C B B C B A B A B B C

Using data set given in the above table assuming the credit rating be the class label attribute. Predict the credit rating for a new customer who is a)short, has red hair and blue eyes.

b)tall, has blond hair and brown eyes. And Predict the eyes when credit rating is B , short height and dark hair.
12) Apply GINI on the data in the table of question number-11 and create decision tree. 13) RainForest is an interesting scalable algorithm for decision tree induction. Develop a scalable

naive Bayesian classification algorithm that requires just a single scan of the entire data set for most databases. Discuss whether such an algorithm can be refined to incorporate boosting to further enhance its classification accuracy. 14) Compare the advantages and disadvantages of eager classification (e.g., decision tree, Bayesian, neural network) versus lazy classification (e.g., k-nearest neighbor, case based reasoning). 15) Design an efficient method that performs effective nave Bayesian classification over an infinite data stream (i.e., you can scan the data stream only once). If we wanted to discover the evolution of such classification schemes (e.g., comparing the classification scheme at this moment with earlier schemes, such as one from a week ago), what modified design would you suggest? 16) What is boosting? State why it may improve the accuracy of decision tree induction. 17) The support vector machine (SVM) is a highly accurate classification method. However, SVM classifiers suffer from slow processing when training with a large set of data tuples. Discuss how to overcome this difficulty and develop a scalable SVM algorithm for efficient SVM classification in large datasets. 18) Write an algorithm for k-nearest-neighbor classification given k and n, the number of attributes describing each tuple. 19) What is associative classification? Why is associative classification able to achieve higher classification accuracy than a classical decision tree method? Explain how associative classification can be used for text document classification. 20) Why is tree pruning useful in decision tree induction? What is a drawback of using a separate set of tuples to evaluate pruning?