CMPUT 695 - Knowledge Discovery and Data Mining
CMPUT 695 Description
It is said that data mining would help us discover the nuggets hidden in the data. What is in fact data mining? There are still many misconceptions about data mining and knowledge discovery from data.
Data mining can help agencies put order and make sense out of their large apparently chaotic data collections and find relevant patterns and significant relationships between information that, on the surface, do not appear connected. It is also vital to scientists and researchers from astronomy to medicine to education toward making discoveries and advance sciences. It is capital to business decision makers toward understanding their customer behaviours. It is paramount for engineers to harness the huge data influx from the different sensors and machinery. It is important for manufacturers to better appreciate and improve their production lines. It is also essential for financiers to assemble knowledge about markets and realize opportunities. Data mining is indeed central to many activities at all levels in our society. Simply put, data mining is the study of efficient ways to discover and extract relevant and actionable knowledge from large collections of data and facts; knowledge that is not necessarily explicitly expressed in the data but intertwined and buried in implied interactions between pieces of evidence and facts hidden in the data; knowledge that is not easily established and acquired.
The goal of CMPUT 695 is to introduce you to the fundamentals of data mining and lift the veil over some misconceptions and capabilities of knowledge discovery today. The course will provide an introduction to knowledge discovery in databases and complex data repositories, and present basic concepts relevant to real data mining applications, as well as reveal important research issues germane to the knowledge discovery domain and advanced mining applications.
Much of the learning in this course is based on the implementation of algorithms in a term project. In the past, many of the projects lead to research publications in international conferences and journals. You will definitely learn considerably and have fun like did other students in this course in previous semesters.
Skills Acquired
Students will understand the fundamental concepts underlying knowledge discovery in data and gain hands-on experience with implementation of some data mining algorithms applied to real world cases.
Course Topics
The course may include the following topics:
- Motivation of Data Mining
(What is data mining and what motivates this research field) - Data Exploration and Analysis
(What are the issues in data analysis and what is the traditional practice) - Association Analysis
(How to find associations in transactional data) - Sequential Pattern Analysis
(How to efficiently discover patterns in sequence data) - Classification and Prediction
(How to categorize data after model learning) - Data Clustering
(How to group data without prior training) - Contrast Sets
(How to compare groups and find the patterns that distinguish them) - Outlier Detection
(How to find exceptions in the data) - Web Mining
(How to mine the heterogeneous and dynamic Web) - Mining Complex Data
(How to mine text, spatial, multimedia and other complex data) - Data Warehousing
(How to organize and aggregate data for decision support)
About the Professor
 |
Osmar Zaïane taught this course at the University of
Alberta since 1999. The course has ever-since evolved and continues to
evolve.
He is currently writing a textbook about the topics
covered in this course.
He is always eager to build confidence and enthusiasm in students
working on research projects. All his supervised graduate students
were selected from his class and successfully encouraged to
publish. He specializes in data mining and is very active in the
research community.
Talk to his graduate students and to students who took this course with him before. |
