Research Task 2 (08/03/15 - 08/16/15)

While you are re-reading the materials from task 1 and learning how to run the OpenViBE with the Mindwave headset, we are going to continue our journey.

The second task includes the study of two important classification techniques in OpenViBE: Support Vector Machine (SVM) and Linear Discriminate Analysis (LDA). They are purely mathematical in nature, so, it is expected to be very challenging! However, don't get frustrated too quick, since most likely, you are not going to develop, modify, or prove those math tools. Instead, you will use OpenViBE to access those tools in your application. So, understand the schemes, learn the vocabularies, be familiar with the mathematical symbols, and grasp the concepts are the most important.

You may need some Linear Algebra to read the math. Feel free to google whatever you don't know. Some matrix operations are not too hard to pick up. Good luck!
     
Classification Algorithm - Support Vector Machine (SVM)

       Start with a series of 3 simple step-by-step tutorials about SVM.

1. SVM - Understanding the math - Part 1 - The margin by Alexandre Kowalczyk.
2. SVM - Understanding the math - Part 2 - Calculate the margin by Alexandre Kowalczyk.
3. SVM - Understanding the math - Part 3 - The optimal hyperplane by Alexandre Kowalczyk.

       Followed by another 2 in-depth videos:

1. SVM Tutorial - Part 1: A YouTube video tutorial on Support Vector Machine.
2. Support Vector Machines (1): Linear SVMs, primal form: Covers the primal form of SVM in a concise way.

Classification Algorithm - Linear Discriminate Analysis (LDA)

       Study the following documents. It may need some repetition in order to grasp the concepts.

1. Pattern recognition systems – Lab 10 Linear Discriminant Analysis.
2. Lecture 10- Linear Discriminant Analysis.
3. A Tutorial on Data Reduction Linear Discriminant Analysis LDA. 
4. Fisher Linear Discriminant Analysis.