I'm using genetic algorithms to train deep reinforcement learning (DRL) agents, similarly to what was done in [this][1] paper. DRL policies are therefore represented by deep neural networks, which map states into deterministic actions. My state space consists of three state variables $v_1, v_2$ and $v_3$. Variable $v_1$ is extremely noisy and seems to be degrading the performance (i.e. the return or discounted cumulative reward) of my RL agent but, for certain reasons, I have to include it. The return is precisely my fitness function. 

Currently, my DNN looks like this:

[![enter image description here][2]][2]

There is only 1 output since the action space is 1-dimensional (one degree of freedom, which is used to control a system).

The DNN tends to overfit more quickly when $v_1$ is present. I'm considering creating a custom NN that looks like this:

[![enter image description here][3]][3]

By doing this I would reduce the complexity of the influence of the variable $v_1$ on the output, since the number of layers between $v_1$ and the output node would be reduced.

I have reasons to believe that the optimal control depends linearly or (something close to linearity) on $v_1$.

Does this make any practical sense and are there are reasons why one should avoid doing this?


  [1]: https://arxiv.org/pdf/1712.06567.pdf
  [2]: https://i.sstatic.net/6h1aC.jpg
  [3]: https://i.sstatic.net/Rp5FX.jpg