Machine Learning - Applications of Machine Learning, Pattern Recognition and Data Mining
The amount and complexity of available data is steadily increasing. In today's business it is paramount to turn this wealth of information into usable knowledge. Machine learning is about developing the required software that automatically analyses data for making predictions, categorizations, and recommendations.
Machine learning algorithms are already an integral part of today's computing systems - for example in customer recommender systems, search engines, financial forecasting, bioinformatics, or biometrical applications - and have reached superhuman performance in some domains. This course will give an introduction to machine learning and its use for pattern recognition and data mining with examples of applications within a wide range of domains.
What you will learn
After course completion you will be able to:
- Recognize and describe possible applications of machine learning for pattern recognition and data mining
- Identify, explain and handle the common pitfalls of machine learning
- Describe and apply techniques for classification and regression
- Use software libraries for solving machine learning problems
- Compare, appraise, and select methods of machine learning for solving specific problems of pattern recognition and data mining
Course content
During the course we will go through some of the essential algorithms used in machine learning for pattern recognition and data mining. We will present the theoretical background as well as give real-world case examples to demonstrate how to use these techniques in practice. You will also get some practical experience in applying these methods on interesting data sets.
We will use the award wining (Gold Prize at Open Source Software World Challenge 2011) Shark machine learning software library (see http://image.diku.dk/shark), developed by co-course director Christian Igel. You can continue using this library on your own projects after the course (the library is open source).
The course will consist of a mix of lectures and hands-on exercises using the Shark software library for machine learning. We will show examples of real world applications as inspiration for you.
The lectures cover the following tentative topics list:
- Foundation of statistical learning
- Classification methods including Linear Discriminant Analysis, K-Nearest Neighbour (KNN), neural networks, and kernel based methods such as support vector machines (SVM)
- Techniques and applications of regression in machine learning
- K-means clustering and mixture modelling
- Dimensionality reduction and visualization techniques such as principal component analysis (PCA)
Participants
This course aims at IT professionals with an education at least at a Bachelor level and/or several years of programming experience.
Machine learning is a fairly mathematical research topic and in order to get the most out of the course you should not be afraid of mathematical symbols. As part of the course, we will give you a quick brush up on some of the mathematical tools we need in order to understand the problem definitions and algorithms (mainly matrices and vectors and some basic statistics).
During the course there will be practical hands-on exercises mainly in C++ and we encourage you to bring a laptop with a C++ compiler preinstalled.
Course dates
5 days, 13 -17 August 2012, 9.00 - 17.00 at the University of Copenhagen, Frederiksberg Campus.
Course directorship
Christian Igel, Professor, Department of Computer Science, University of Copenhagen.
Kim Steenstrup Pedersen, Associate Professor, Department of Computer Science, University of Copenhagen.
Course fee
DKK 18,500 / EUR 2,500 (excl. Danish VAT 25%) price includes teaching, course materials and all meals during the course. If you are a member of Finansforbundet, please read more
