PyAI

A machine learning / artifial intelligence framework written for python designed to be easy to use and felxible. It is based mostly on scikit-learn and numpy, but incorperates many open-source and personal machine learning libraries.

Requirements

• Python 2.x
• numpy
• scipy
• scikit-learn
• matplotlib
• decorator

Installation Instruction

1. python setup.py build
2. sudo python setup.py install
3. In python: import PyAI
4. Test by typing: PyAI.test()

Usage

The main object in the library is the Brain class (PyAI.Brain). With it you access all of the features in the framework.

brain = PyAI.Brain(x_data=data, y_labels=labels, y_data=reg_data)

This brain object has 2 modes of operation: classification and regression.

• If you wish to perform classification (discrete) prediction, use the y_labels attribute
• If you wish to perform regression (continuous) prediction, use the y_data attribute
• Or you can also provide both

Then, you must initialize one of the algorithms available by performing:

brain.init_XXX()
# For example
brain.init_clustering(n_clusters=5)

Currently, the available algorithms are

• clustering
• neighbors
• svm
• gmm
• naive_bayes

Then you can apply any number of prediction methods in order to predict using the models

brain.predict_xxx_yyy
# For example
brain.predict_cluster_labels(test_data)
brain.predict_svm_data(test_data)

The xxx must match on of the algorithms that you have initialized

The yyy can either be 'labels' or 'data' for classification and regression respectively