SI - Pattern Recognition

Table-of-contents

• Dependencies
• Features extraction
• To perform the practise

More description is given in the subsections.

Dependencies

The following practise has been tested with Ubuntu 14.04.

In order to use the Ipython Notebook, the following dependencies are needed:

• IPython - sudo apt-get install ipython
• IPython notebook - sudo apt-get install ipython-notebook
• Numpy - sudo apt-get install python-numpy
• Scipy - sudo apt-get install python-scipy
• Matplotlib - sudo apt-get install python-matplotlib
• Mpld3 - sudo pip install mpld3
• Plotly - sudo pip install plotly
• Scikit-learn - sudo apt-get install python-sklearn

We strongly recommend to use a Linux environment to perform this practise.

Features extraction

The following module of the framework will be studied:

Synthetic data will be generated. An implementation of Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA) will be implemented. A comparison with the scikit-learn implementation can be performed.

Principal components analysis (PCA)

Insights about PCA can be found in this nice article here by D. Sidibe.

Linear discriminant analysis (LDA)

Insights about LDA can be found in this article here.

To perform the practise

Assignment procedure

In order to perform the practise, you will have to fork the current project. To do so,

• Fork the current project by click on the Fork icon ,
• Select your GitHub profile if necessary,
• Clone the repository ,
• Solve the practise by executing the Ipython notebook,
• Commit & push your changes in your own repository,
• Make a pull request.

Execute the Ipython notebook

Enter the following command in a terminal ipython notebook.

This command should run the server locally via your default web browser and you will be able to play with the notebook.

If you are just curious to see what the ipython notebook look like, you can view it there.

Enjoy!!!