Syllabus Data mining Introduction to Data Mining

Syllabus Data mining

1. 
   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 
     
2. 
   Data Warehouse and OLAP 
   
   • 
     Data Warehouse and DBMS 
     
   • 
     Multidimensional data model 
     
   • 
     OLAP operations 
     
   • 
     Example: loan data set 
     
3. 
   Data preprocessing 
   
   • 
     Data cleaning 
     
   • 
     Data transformation 
     
   • 
     Data reduction 
     
   • 
     Discretization and generating concept hierarchies 
     
   • 
     Installing Weka 3 Data Mining System 
     
   • 
     Experiments with Weka - filters, discretization 
     
4. 
   Data mining knowledge representation 
   
   • 
     Task relevant data 
     
   • 
     Background knowledge 
     
   • 
     Interestingness measures 
     
   • 
     Representing input data and output knowledge 
     
   • 
     Visualization techniques 
     
   • 
     Experiments with Weka - visualization 
     
5. 
   Attribute-oriented analysis 
   
   • 
     Attribute generalization 
     
   • 
     Attribute relevance 
     
   • 
     Class comparison 
     
   • 
     Statistical measures 
     
   • 
     Experiments with Weka - using filters and statistics 
     
6. 
   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 
     
7. 
   Data mining algorithms: Classification 
   
   • 
     Basic learning/mining tasks 
     
   • 
     Inferring rudimentary rules: 1R algorithm 
     
   • 
     Decision trees 
     
   • 
     Covering rules 
     
   • 
     Experiments with Weka - decision trees, rules 
     
8. 
   Data mining algorithms: Prediction 
   
   • 
     The prediction task 
     
   • 
     Statistical (Bayesian) classification 
     
   • 
     Bayesian networks 
     
   • 
     Instance-based methods (nearest neighbor) 
     
   • 
     Linear models 
     
   • 
     Experiments with Weka - Prediction 
     
9. 
   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 
     
10. 
    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. 
      
11. 
    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 
      
12. 
    Advanced techniques, Data Mining software and applications 
    
    • 
      Text mining: extracting attributes (keywords), structural approaches (parsing, soft parsing). 
      
    • 
      Bayesian approach to classifying text 
      
    • 
      Web mining: classifying web pages, extracting knowledge from the web 
      
    • 
      Data Mining software and applications 
      

Hodnotenie a práca počas semestra

Na začiatku cvičenia bude krátka 5 minútovka. Výsledky 5 minútoviek tvoria 20% hodnotenia výslednej známky. Výsledok 5minútoviek zistíte na konci semestra.

Toto je nutná podmienka pre splnenie zápočtu.

Skúška bude pozostávať z testu. Teda už sa nebudú robiť žiadne projekty. Na teste dostanete dataset a zadanie.