This repository contains the application of five common machine learning models to forecast prices of the stocks from the UCI Dow Jones Index Repository. The models are the following :
- Linear SVM
- RBF SVM
- Linear Regression
- Ridge Regression
- One Hidden Layer MLP
To see the process of finding the best model from all the types of models, execute the following file in R:
find_best_model.R
You can check for each individual method what are the best parameters for them executing:
best_svm.R
best_svm_rbf.R
best_linear_regression.R
best_ridge_regression.R
best_mlp.R
For each one of these also a visualization of the difference between the forecast and the real price will appear. All models are generated in regression_models.R
.