Support vector machines for face recognition:

Face recognition is a learning problem that has recently received a lot of attention. Support Vector Machines (SVM) are becoming very popular in the machine learning community as a technique for tackling high-dimensional problems. Please implement the SVM algorithm by yourself. The experimental dataset can be found at "http://www.cl.cam.ac.uk/research/dtg/attarchive/facedatabase.html". 

The required steps:
1) Implement the linear SVM. You can use library (e.g. MATLAB "quadprog" function) to solve the quadratic programming problem. For the rest implementation, you need write your own code.

2) Face recognition is a multi-class classification problem. You can use either one vs rest or one vs one strategy.

3) You can use half data for training and the rest half data for testing. After that, you can switch training and testing data. The final result is the average of two accuracy values.