darker66678/hyperml

Optimizing ML hyperparameter problem using heuristic algorithm

HyperML

Optimizing ML hyperparameter problem using heuristic algorithm

Updated log

• Updated log

Support algorithms

Heuristic

• PSO
• VOA

Machine Learning

Classification

• SVM
• MLP
• KNN
• Random Forest
• Adaboost

Regression

• SVM
• MLP
• KNN
• Random Forest
• Adaboost

Scoring

All the scores

How to start

Requirements

Python 3.8

In anaconda, conda create --name test python=3.8 can create a new env

pip install -r requirements.txt install all packages

Task

Hyperparameter

python ./hyperml/main.py hyper start to tune the hyperparameters with heuristic

Please refer following information on parameter config for your experiment

hyper_parser.add_argument("-m","--model", help="KNN,MLP,SVM,RF,ADA,XGBoost", default="ADA")
hyper_parser.add_argument("-a","--algo", help="PSO,VOA,RANDOM", default="RANDOM")
hyper_parser.add_argument("-s","--scoring", help="cls: accuracy, f1, recall, precision, specificity; reg: r2, neg_mean_absolute_error, neg_mean_squared_error", default="r2")
hyper_parser.add_argument("-k","--k_fold", help="set k value , need to >1", default=3, type=int)
hyper_parser.add_argument("-c","--confusion_m", help="Do you need to gernerate the confusion_matrix?(False or True)", default=False, type=bool

Clustering

python ./hyperml/main.py cluster

Feature Selection

python ./hyperml/main.py feat_select

The results will be saved in ./results

Video

Tutorial