sciblox 0.2.5

Making data science and machine learning in Python easier.

sciblox was designed to make data science and machine learning easier.

It features similar modules to R’s CARET, and draws inspiration from many other packages.

To use it:

from sciblox import * %matplotlib notebook

What’s included?

1. CARET like preprocessing (scaling, normalising, dummify, outlier removal, unskew, etc)

2. Processor modules - you can fit onto testing data!

3. LightGBM and RandomForest support - later will add more

4. More analysis methods: outlier detection, skewness methods, auto unskewing etc

5. 3D automatic plots in NEW cool analyse function!

6. BPCA, MICE, KNN, Boosting, Forest, SVD imputation

7. Easy sequential text mining and automatic text mining

8. Jupyter Notebooks integrated

What’s in construction?

1. More machine learning libraries - Extra Trees, Neural Networks, SVM etc

2. Advanced data and text mining

3. CV, Auto machine learning

4. Multiprocessing support

Demonstration

If you want to see sciblox in action, please visit https://danielhanchen.github.io/

GITHUB: https://github.com/danielhanchen/sciblox