--

Update1:

I am not trying to define different categories of policies, If you want to see how well each category of policy is clearly defined, you might just find that neither on-policy nor offline policy definations are correct.

I was only trying to pointing some of the fundamental differences ( WHICH IS BEHAVIOURAL POLICY AND UPDATE POLICY and how they play out in different setting ) which I felt is most important. I clearly mentioned the same in the beginning of the blog.

I conveniently choose to explain off-policy with ERB because its easy for my reader to understand the " DIfference". Regardless of very harsh beginning of your comment, I will try to re-emphasise the point by editing the blog.

Update2

You are welcome !

--

--

Kowshik chilamkurthy
Kowshik chilamkurthy

Written by Kowshik chilamkurthy

RL | ML | ALGO TRADING | TRANSPORTATION | GAME THEORY

Responses (1)