Using what you have learnt from Submissions 1 and 2, implement a trading strategy using
machine learning. We recommend that students focus on classification – for example: trying to
forecast if a stock will move up and down, above some threshold such as the 90-day standard
1 Decide on an algorithm or group of algorithms (for example, ensemble techniques).
2 Fit the model.
3 Show that it works out of sample, and use appropriate cross-validation techniques.
4 Provide the following performance metrics:
(a) ROC curves,
(b) Confusion Matrix,
(c) Precision, Recall, F1-Score, Accuracy, and AUC.
5 Analysis of metrics and report.
Create a fund factsheet for your new investment strategy. Have a look at examples of popular
funds found online and create a fact sheet with all the bells and whistles. It must at a minimum
include (Pyfolio can be used):
1 Maximum Drawdown
2 Annualized Returns
3 Sharpe Ratio
4 Plot the Equity Curve