Report of Multi-variable prediction of MSFT stock price prediction. Compared between 3 models -- Logistic Regression, Gradient Boosting Machines(XGBoost), Neural Networks (a Multi-Layer Perceptron for classification).
Conclusion: With the complicate Market data, complex models(XGBoost, MLP) failed to out-perform the simple Logistic Regression.
Click the badge above to view the full notebook with all interactive charts and data outputs.