by Gabriel Turinici, Pierre Brugiere
Continue on SelfMadeQuant
Suggested foundations for financial machine learning and financial time series. Specialist prerequisites vary by paper.
Preserve time order and avoid future information in training.
Read the fitted relationship and its residuals.
Check whether a relationship can be estimated consistently over time.
Account for dependence and overlapping observations.
Examine which information contributes to predictions.
Study long-run relationships without confusing them with correlation.
Choose a reproducible strategy, specify the trading rule, and test what remains after costs and estimation uncertainty.