Advanced Pattern Recognition Course for Industry
Are you working on practical classification problems but you miss the necessary knowledge or skills concerning the use of good statistical techniques to analyse and learn from your data?
- Are you overwhelmed by the abundance of available learning methods? Do you miss clarity?
- Do you want to learn practical approaches to analyze and learn patterns from the data?
- Are you interested in combining statistical learning tools with the image processing techniques?
- Are you looking for fundamental understanding and hands-on experience of pattern recognition methods?
Just imagine for a moment that you had knowledge, understanding and a toolbox of practical methods to make it a reality. It would be great, wouldn’t it? Well, it is all within your reach!
Our 5-day Advanced Pattern Recognition Course will help you to extend your competence and practical capabilities. The course is given by Delft Pattern Recognition Laboratory at TU Delft, in cooperation with the Dutch Neural Network Foundation (SNN).
This course is unique in the world, since it offers our broad experience and expertise in the research of applied and fundamental pattern recognition. Moreover, it provides a balance between theory (lectures) and practical exercises. It has constantly been refreshed, improved and updated over the last eight years. The course is using the PRTools toolbox developed at TU Delft which provides great flexibility and can be adopted in industrial circumstances, as many of our participants have already done. The course is based on:
- working in small groups (4 to 10 participants), ensure that you benefit from informal discussions as well as personal assistance
- intensive hands-on experience with advanced pattern recognition tools (5 hours/day), we create a friendly but challenging atmosphere to optimally stimulate your learning process. But we will also match your pace individually through our assistance during the labs
- exercises with real world data and applications, you are most welcome to bring your own data to work with
- two lectures per day (60-75 min. each), giving fundamental understanding
You may find more details in the preliminary program.
This includes :
- lectures and labs
- personal assistance
- all the materials
- lunches and refreshments
Please note that accommodation is not included (see below).
Teaching and laboratory staff
|Dr.Ir. R.P.W. Duin
Bob Duin is internationally recognised as one of the key experts in the field of pattern recognition, being known from his practical and well-founded approaches.
|Dr. Ela Pekalska,My expertise lies in pattern recognition with the emphasis on representation and generalisation. My interests include kernel methods, proximity-based learning, novelty detection, combining paradigms, dimensionality reduction and Gaussian processes. I am also keen on understanding brain, neuroscience and psychology.
|Dr. David Tax,David Tax has his main expertise in the field of one-class classification (or outlier detection), but also works on kernel methods, ROC analysis and Multi-Instance Learning.
Who should attend
Researchers and system designers working in :
- industrial research and development
- pharmeceutical/medical research and development
- intelligent sensors design
- traffic control/monitoring, telematics
and all applied research organizations needing tools for advanced data analysis and/or preparing applications in need of intelligent handling of sensory data.
Previous attendants include researchers from :
- Advanced Medical Diagnostics Holding SA (AMDH) – Belgium
- Autoliv, automotive safety systems – Sweden
- Carathian Tech Research (CTR) – Austria
- CIRA, Italian Aerospace Research Center – Italy
- De Beers Consolidated Mines – South Africa
- DoD, Department of Defence – United States
- Noldus, systems for observational data – The Netherlands
- Philips – The Netherlands
- TNO – The Netherlands
Some experience with computer programming and a basic familiarity with linear algebra and statistics. The courses are based on Matlab tools. Some knowledge of Matlab is advantageous, but not necessary.
We will provide a workstation running Matlab, a disk with the PRTools and perClass Matlab toolboxes, exercises and data, the lecture sheets.
Registration should be made not later than 17 days before the start of the course. Payment of the registration fee should be made upon receipt of the invoice and should be received 10 days before the start of the course.
Cancellation is possible until 7 days before the start of the course. In this case, 50% of the original fee will be refunded. After this date, refunding is not possible.
Pattern Recognition Laboratory
Faculty of Electrical Engineering, Mathematics and Computer Science
Delft University of Technology
Mekelweg 4, Delft
The building is on 15 minutes walking distance from the center of Delft, one of the most beautiful cities of The Netherlands. Besides Delft, the cities of Rotterdam and Amsterdam are close and worth a visit.
In order to reserve a hotel in Delft you can visit, for example, www.delfthotels.nl
If you have any questions, please send us an e-mail.