CS580 - Data Mining, Summer 2023 (online)

Course Information

Course title: Data Mining
Course number: CS 580
Course description: Data Mining studies algorithms and computational paradigms that allow computers to find patterns and regularities in data, perform prediction and forecasting, and generally improve their performance through interaction with data. It is currently regarded as the key element of a more general process called Knowledge Discovery that deals with extracting useful knowledge from raw data. The knowledge discovery process includes data selection, cleaning, coding, using different statistical and machine learning techniques, and visualization of the generated structures. The course will cover all these issues and will illustrate the whole process by examples. Special emphasis will be given to the Machine Learning methods as they provide the real knowledge discovery tools. Related technologies, as data warehousing and on-line analytical processing (OLAP) will be also discussed. The students will use recent Data Mining software. Enrollment in this course is limited to 15 students.
Course dates: May 30, 2023 - Jul 24, 2023

Instructor Information

Name: Zdravko Markov
Email: markovz@ccsu.edu
URL: http://www.cs.ccsu.edu/~markov/
Office location: 30307 Maria Sanford Hall, CCSU
Phone: (860) 832-2711

Textbook

Required reading: Ian H. Witten, Eibe Frank, and Mark A. Hall, Christopher Pal. Data Mining: Practical Machine Learning Tools and Techniques (Fourth Edition), Morgan Kaufmann, January 2017, ISBN 978-0-12-804291-5.
Required Software: Weka 3: Data Mining System with Free Open Source Machine Learning Software in Java. Available at http://www.cs.waikato.ac.nz/~ml/weka/index.html

Course Goals

Introduce students to the basic concepts and techniques of Data Mining.
Develop skills of using recent data mining software for solving practical problems.
Gain experience of doing independent study and research.

Grading Policies

Grading will be based on eight assignments (75%), one quiz (10%) and class participation through three scheduled discussions (15%). The maximum course total is 1000 points. The letter grade will be determined by the following grading scale:

A	A-	B+	B	B-	C+	C	C-	D+	D	D-	F
950-1000	900-940	870-890	840-860	800-830	770-790	740-760	700-730	670-690	640-660	600-630	0-590

Late assignments will be marked one letter grade down for each 3 days they are late. It is expected that all students will conduct themselves in an honest manner and NEVER claim work which is not their own. Violating this policy will result in a substantial grade penalty or a final grade of F.

Course Content (12 units)

Introduction to Data Mining

What is data mining?
Related technologies - Machine Learning, DBMS, OLAP, Statistics
Data Mining Goals
Stages of the Data Mining Process
Data Mining Techniques
Knowledge Representation Methods
Applications
Example: weather data

Data Warehouse and OLAP

Data Warehouse and DBMS
Multidimensional data model
OLAP operations
Example: loan data set

Data preprocessing

Data cleaning
Data transformation
Data reduction
Discretization and generating concept hierarchies
Installing Weka 3 Data Mining System
Experiments with Weka - filters, discretization

Data mining knowledge representation

Task relevant data
Background knowledge
Interestingness measures
Representing input data and output knowledge
Visualization techniques
Experiments with Weka - visualization

Attribute-oriented analysis

Attribute generalization
Attribute relevance
Class comparison
Statistical measures
Experiments with Weka - using filters and statistics

Data mining algorithms: Association rules

Motivation and terminology
Example: mining weather data
Basic idea: item sets
Generating item sets and rules efficiently
Correlation analysis
Experiments with Weka - mining association rules

Data mining algorithms: Classification

Basic learning/mining tasks
Inferring rudimentary rules: 1R algorithm
Decision trees
Covering rules
Experiments with Weka - decision trees, rules

Data mining algorithms: Prediction

The prediction task
Statistical (Bayesian) classification
Bayesian networks
Instance-based methods (nearest neighbor)
Linear models
Experiments with Weka - Prediction

Evaluating what's been learned

Basic issues
Training and testing
Estimating classifier accuracy (holdout, cross-validation, leave-one-out)
Combining multiple models (bagging, boosting, stacking)
Minimum Description Length Principle (MLD)
Experiments with Weka - training and testing

Mining real data

Preprocessing data from a real medical domain (310 patients with Hepatitis C).
Applying various data mining techniques to create a comprehensive and accurate model of the data.

Clustering

Basic issues in clustering
First conceptual clustering system: Cluster/2
Partitioning methods: k-means, expectation maximization (EM)
Hierarchical methods: distance-based agglomerative and divisible clustering
Conceptual clustering: Cobweb
Experiments with Weka - k-means, EM, Cobweb

Text and Web Mining

Basics of Information Retrieval
TFIDF representation of text documents
Text classification