Quantopian zipline trading algorithm parameter optimization with Spearmint Bayesian Optimizer - Part 5 Conclusions and Next Steps

Using the open source spearmint optimizer to select optimal tuning parameters for a quantopian zipline trading algorithm provides the quant trader with a powerful tool.  It is far more efficient to use an optimizer like spearmint than to perform a grid search.
 
There is a comercial version of spearmint available from whetlab which I have had a chance to preview.  It has a lot more features and I would recomend quants explore this if they wish to pursue this approach.
 
In this series, I used a cloud based Debian linux virtual machine of modest size (2 CPU, 4 GB of memory) and obtained a feasable solution in under two hours.   More complex algorithm with longer time frames would take more resources of course.  A major factor in convergence time is the sensitivity of the tuning constants.
 
If you would like me to build out an equivalent image on the cloud for you, please ContactMe .
 
Back to Part 4 or back to Part 1 .