Author: Yuzhen Liu
Started From 2019.3.22
Light-weighted Statistic ML implementations in C++. Algorithms included are linear regression, Logistic Regression Classifier, Sofrmax Classifier, C4.5 Decision Tree, Random Forest, GBDT, FM, Naive Bayes Classifier, SVM.
Requeirs C++ algebra library armadillo, installation of armadillo is as follows:
install denpencies first (Tested for Ubuntu 16.04):
sudo apt-get install libopenblas-dev
sudo apt-get install liblapack-dev
sudo apt-get install libarpack2-dev
sudo apt-get install libsuperlu-dev
download armadillo (armadillo-9.300.2 tested) as xxx.tar, cd and build it.
cd armadillo-9.300.2
cmake .
make
sudo make install
clone and build Arma_ML directly in place
git clone https://github.com/codestorm04/Arma_ML.git
cd Arma_ML
make
or build and install
make install
examples/ are the usage demos of each modules, reference to README.md
- Build strategies: Reference
- decision tree pruning
- models: knn, LDA, PCA, MDS, k-means, FFM
- params setting
- normalization: L1 L2
- optimazers other than SGD
- model saver / loader
- oprimizers, metrics, console visualization module