Alternative parameter tuning


I would like to tune parameters for a backtest function. How can I tune parameters NOT hyperparamters without using brute force?

Thank you for your query, Emma.



An effective method for parameter optimization involves a randomized search. This technique selects parameter combinations randomly within defined ranges. Unlike exhaustively examining every potential combination, which can be particularly time-consuming in complex, high-dimensional parameter spaces, this approach randomly pinpoints values within the predetermined ranges for each parameter.



The subsequent steps remain the same, involving the division of the dataset into training and testing subsets. Then identifying the optimal parameters within the training dataset via random search and subsequently evaluating these parameters in the test dataset.



We've also delved into parameter optimization using Machine Learning (clustering) in the section advanced alpha mining of the Trading Alpha course.



Quantra by QuantInsti | Courses on Algorithmic and Quantitative Trading





Thanks!