Run function- Can we move it to the 21 century?

AI & Machine learning
1

Course Name: Deep Reinforcement Learning in Trading, 

1. The run function, just runs- endlessly- and crashes. Can it be made to stop running, without manually crashing it, while preserving the results from 24 hrs run? At this time, the 'run' function just runs blind crazy: no stopping/starting it, while preserving the `experience` in training. 

2. Can we control what it shows and where it makes its trades? Like on a chart? 

Thanks!

2

Hi Antoni,



The deep reinforcement learning algorithm does take time to run. To reduce it, you could add a stopping criteria, or a save function, which saves the 'experience' after a certain number of trades. You can also try to set a maximum number of training episodes, or you can try reducing the number of iterations or setting the batch size to a lower number.



Once the model has been trained and takes trades, you can use python to plot the returns as well as the time it took a certain trade. 



Hope this helps.