Python Snippets for Machine Learning Learning

This repository covers some developed/in development snippets used for machine learning in Python, such as: CSV to ARFF, CSV percentage split.

TECHNOLOGIES & LIBRARIES USED

1. Python 2.7

SNIPPETS

1. CSV percentage split (supervised learning): insert a CSV file with semicolon delimitators and split proportionally to each true label, exporting two files - one for training (normally 80%) and another for testing (normally 20%).

HOW TO USE

1. Set permissions to split-supervised-learning.py to run:

$ chmod +x split-supervised-learning.py

2. Insert the CSV file to be splitted inside /data/raw.

3. Make sure the delimitators for each value is a comma and the header titles are between quotes.

4. Run python script with the true label name, the file name from /data/raw (without .csv extension), the training-rate and the testing rate. Examples:

$ ./split-supervised-learning.py lettr letters 80 20
$ ./split-supervised-learning.py quality wine 90 10

REFERENCES

1. Lichman, M. (2013). UCI Machine Learning Repository. Irvine, CA: University of California, School of Information and Computer Science.

2. Letter Recognition Data Set.

3. P. Cortez, A. Cerdeira, F. Almeida, T. Matos and J. Reis. Modeling wine preferences by data mining from physicochemical properties. In Decision Support Systems, Elsevier, 47(4):547-553, 2009.

COLABORATORS

KAWASAKI, Davi // davishinjik [at] gmail.com

FLAUSINO, Matheus // matheus.negocio [at] gmail.com

CONTACT & FEEDBACKS

Feel free to contact or pull request me to any relevant updates you may enquire:

KAWASAKI, Davi // davishinjik [at] gmail.com