Introduction: This course is an elective course for master students majoring in Computer science and technology or related. It mainly introduces the basic concepts and algorithms of data mining, and discusses and explores the academic frontier of data mining. The purpose is to provide knowledge reserve for students' study and research.

---

         

Course leader (lecturer): Wang Qian, female, born in 1987, associate professor, Doctor, doctoral supervisor. Her main research interests are data mining, machine learning and network security. From March 2015 to March 2016, she was engaged in the research work related to computer science and technology in the University of Hull, UK. Email: wangqianysu@163.com

 

       

Lecturer: He Haitao, female, born in 1968, professor, doctor, doctoral supervisor. Her main research interests are artificial intelligence and data mining. From September 2011 to March 2012, she was engaged in the research work related to computer science and technology in the University of Warwick, UK. Email: haitao@ysu.edu.cn

    

Lecturer: He Hongdou, male, born in 1991, lecturer, Doctor, master supervisor. His main research interests are data mining and machine learning. Email: hhd@ysu.edu.cn                    

 

---

I. Basic requirements of this course

1. To master the important concepts and tasks of data mining, data warehouse and OLAP analysis technology, and classic algorithms of data mining.

2. To master the specific operation process of data mining, be able to use simple data mining algorithms with certain data analysis and processing capabilities, and to solve practical problems.

3. To master the evaluation indicators of algorithms, and be able to evaluate the performance of data mining algorithms from multiple angles.

4. To understand the current research trends and research hotspots of data mining, and to understand the development direction and dynamics of data mining technology.

II. The basic content of the course

Section 1: Theoretical class

1. Introduction (2 credit hours)

Teaching objective: To understand the concepts related to data mining and the functions of data mining.

Main contents:

(1) Basic concepts of data mining

(2) What kind of data is mined

(3) The functions of data mining

2. Data preprocessing (2 credit hours)

Teaching objective: To master the definition and method of data cleaning, data integration and transformation, and data specification. To understand the process and method of data preprocessing and the concept and method of discretization and concept stratification.

Main contents:

(1) Data cleaning methods

(2) Definition and method of data integration and transformation

(3) Definition and method of data reduction

(4) The generation of discretization and concept stratification

3. Overview of Data warehouse and OLAP Technology (2 credit hours)

Teaching objective: To master the data model of data warehouse, the system structure of data warehouse and OLAP operation on multidimensional data model.

Main contents:

(1) Concept of data warehouse

(2) Multidimensional data model, star and snowflake database model

(3) OLAP operation on multidimensional data model

(4) System structure of data warehouse

4. Data cube calculation and data generalization (3 credit hours)

Teaching objective: To master concept description, data generalization definition and summary-based characterization, to understand attribute correlation analysis and to distinguish different classes.

Main contents:

(1) An effective method of data cube calculation

(2) Data generalization and summary-based characterization

(3) Analytical characterization: attribute correlation analysis

(4) Mining class comparison: distinguish different classes

5. Mining frequent patterns, associations, and correlations (2 credit hours)

Teaching objective: To master the Apriori mining algorithm of association rules and the mining process.

Main contents:

(1) Basic concepts of association rule mining

(2) Efficient and scalable frequent itemset mining method

(3) Multi-layer association rules are mined by transaction database

(4) From association rules to correlation analysis

6. Classification and Prediction (3.5 credit hours)

Teaching objective: To master the concepts of classification and prediction, to master the data preparation, and classification algorithms such as decision trees.

Main contents:

(1) Concepts and steps of classification and prediction

(2) Data preparation for classification and prediction

(3) Decision tree induction classification algorithm

(4) Bayesian classification algorithm

7. Cluster Analysis (2.5 credit hours)

Teaching objective: To master the concept of cluster analysis, understand the difference between clustering and classification, and master K-means and other clustering algorithms.

Main contents:

(1) The concept of cluster analysis

(2) Data types in cluster analysis

(3) Classification of main clustering methods

(4) Partition based method

Section 2: Discussion class

1. Teaching objective

The discussion class is a discussion on a certain issue, with the purpose of deepening students' understanding of data mining theoretical knowledge, and cultivating students' ability of independent thinking, independent analysis, problem solving and oral expression.

2. Main content

(1) Illustrate the data mining mode needed in daily life.

(2) Examples of data warehouse model and data preprocessing methods.

(3) Evaluation of classical data mining algorithms and comparison of advantages and disadvantages.

(4) Research hotspots and development trends of data mining.