Back to Community
Data Snooping Bias Reminder

Came across this while reading Ernie Chan's "Quantitative Trading". As someone starting out on the road of Quant trading I have been guilty of doing this and i think it's worth reiteration to the community here where there are others starting out like myself.

"In general, you should eliminate as many conditions, constraints, and parameters as possible as long as there is no significant decrease in performance in the test set, even though it may decrease performance on the training set. (But you should not add conditions and parameters, or adjust the parameter values, so as to improve performance on the test set: If you do, you have effectively used the test set as your training set and possibly reintroduce data-snooping bias to your model.)"