Course talk:CPSC522/Genetic Algorithms

Overall, this article is well-written and interesting, especially for someone who has no prior knowledge in this. As the other critique mentions, the applications part could be fleshed out a bit more as the article is a little too concise. Here are a couple of minor comments:

Critique

1) It would be great if you can number the sub-headings under Example. Maybe: Step 1, Step 2 etc.? Initially, I didn't realize that the subheadings were going to sequentially build the example.

2) "For example, if there were 12 chromosomes, there would be six remaining after halving. Then, if a chromosome had rank 1, then its probability would be 6/21." Could you please further explain how a rank 1 chromosome has a probability of 6/21?

3) Not quite sure if Evolutionary Reinforcement Learning is the most important application of Genetic Algorithms. Why are we looking into this in particular while the other applications are combined in a small paragraph below? Some further motivation for why we're specifically looking into Evolutionary Reinforcement Learning would be appreciated.

4) Also, Evolutionary Reinforcement Learning kind of shows up in the article out of nowhere with little to no introduction, immediately after the previous example has been discussed. A little introduction (where you also possibly highlight the significance of this particular application) would really help.

2) A citation/reference for evolutionary reinforcement learning would be appreciated for anyone who wants to read more about this.

Minor edits 1) *In the ERL model, agents interact with an environment containing food, predators, and other objects.

Very interesting page! I love this algorithm and the page is presented in a way that is very clear.

• I would recommend adding some lines/bars to the pseudocode, it is a bit hard to read.

• There should be gifs/videos on these algorithms right? I know there are a handful of amazing presentations, and I would love some mention of those!

• while I enjoyed the read, it does seem a bit short to me, but I would love some more up-to-date papers that are more technical.

All in all, I don't think there are many errors on the page. I understand that Genetic Algorithms may not be very technical, but I would really appreciate it if there were more interesting discussions on the future of these algorithms are and what is the SOTA model. Maybe just a literature review on the various fun/famous applications?

Regardless I enjoyed my read!