Sample test data are published at UCI.
- Some nice list of Python packages for Python coding. Another its link is here.
- Standard Python packages include
pandas, numpy, scipy, matplotlib, scikit-learn, seaborn
- H2O for R and Python
- The programs run on Java SDK.
- For Python version, check out the installation instruction here. Basic introduction to its useage is here. If there is a conflict between Linux OS system
pip
and Anaconda'spip
, and you wish to install it on Anaconda envrionment, then use/root/anaconda/bin/pip ***
instead ofpip install ***
- Some of its powerful tools are Deep Learning and GBM. Its deep learning so far primarily focuses on the model of multi-layer, feedforward neural networks for predictive modeling.
- Other useful packages include
seaborn
for data visualization.Orange
for graphic modeling instead of source coding.Theano
for deep learning and tensor calculation.numba
for JIT (just-in-time compilation) when it involves (especially looping) many generic Python codes and numpy.