Posts tagged 'data-leakage'
Define timing, data splits, retraining, and cost rules for a fair comparison of logistic regression and baseline strategies, before testing actual performance.
Separate signal, execution, return, and label timing. Use synthetic unit tests to check information leakage and preprocessing boundaries.
Organize LLM-assisted quant research around questions, evidence, code checks, timing, and experiment records, with human review and documented failures.
Examine noise, changing markets, repeated searches, leakage, costs, and selection bias, and define ways to challenge promising financial ML results.
Design features, labels, time splits, and monthly retraining for a logistic-regression baseline, keeping predictive accuracy separate from investment returns.
Audit historical data availability through adjusted prices, delistings, filing dates, revisions, and time zones, with an example of delayed closing-price signals.
Forge Fate