/rusmalai-ncbs

Repository of all the codes using during the initial phase of RUSMALAI journal club in NCBS.

Primary LanguageJupyter NotebookMIT LicenseMIT

RUSMALAI-NCBS

Repository of all the codes using during the initial phase of RUSMALAI journal club in NCBS.

RUSMALAI stands for Reinforcement Unsupervised MAchine Learning and Artificial Intelligence. RUSMALAI is a weekly journal club for machine learning in NCBS (National Centre for Biological Sciences, Bangalore). The codes in this repository were written as a part of the initial phase of the Journal Club, when the objective was to attain sufficient understanding of core concepts. You can find more about the objective in our manifesto.

Topics covered (will be continuously updated as the journal club progresses)

  1. Perceptrons - a simple example of a perceptron based classification of two features on the iris dataset.
  2. Backpropagation learning - iris dataset
  3. XOR problem - Multi layer perceptrons - iris dataset
  4. Hopfield networks - Digit pattern completion
  5. Beginner's mnist - Using tensorflow
  6. Deep mnist - Using tensorflow

Requirements (to get the ipython notebooks working)

The scripts is written in python. It requires the following packages:

  1. scikit-learn
  2. scipy
  3. numpy
  4. matplotlib
  5. ipython
  6. jupyter

These can manually installed by using pip or by using the requirements file given in the repository using the following command:

[sudo] pip install -r requirements.txt