21–27 of 27 posts
"A Field Guide to Genetic Programming" ( https://www.amazon.com/Field-Guide-Genetic-Programming/dp/14... ) is one of the better books I've read on the matter (but I've yet to read anything by Koza). W. Langdon's webpage ( http://www0.cs.ucl.ac.uk/staff/W.Langdon/ ) has a lot of great information.
Evolution Strategies is not an example of Genetic Programming, and that book by Langdon doesn't cover anything directly relevant to the linked article.
I feel like this is more an argument against the efficiency of flat (non-hierarchical, not model-based) Deep Reinforcement Learning than an argument for Genetic/Evolutionary Algorithms. As in, if you're not solving a task more efficiently than Evolutionary Algorithms or more difficult than they can handle, you're not doing something right. Similar to how Deep PCA and Deep Random Forests blew ConvNets out of the water…
Deep PCA: https://arxiv.org/abs/1404.3606 Deep Forest: https://arxiv.org/abs/1702.08835
http://www.cv-foundation.org/openaccess/content_iccv_2015/pa...
That's Deep Neural Decision Forests; they benchmark against MNIST and obtain state of the art results.
The other paper is about improving decision forests, although it also uses MNIST as a benchmark.
Apologies if I'm wrong and you did mean the Zhou & Feng paper.
Reading the code; I dont get get: 1. How to model the different actions 2. Tie future awards to action at this timestep... Can anyone help with better nitty-gritty explanation ?
The example code inline in the article just illustrates the basic idea of Evolution Strategies (ES), not their new work in applying ES. The behavior of agents is determined by a "policy function". This function takes in inputs (e.g. what the agent sees) and outputs actions (e.g. what the agent does). The policy function has a set of internal parameters that determines the precise mapping from inputs to outputs. In th…
Earlier quoted context omitted.
Deep PCA: https://arxiv.org/abs/1404.3606 Deep Forest: https://arxiv.org/abs/1702.08835
I think this is the Deep Forest paper you meant: http://www.cv-foundation.org/openaccess/content_iccv_2015/pa... That's Deep Neural Decision Forests; they benchmark against MNIST and obtain state of the art results. The other paper is about improving decision forests, although it also uses MNIST as a benchmark. Apologies if I'm wrong and you did mean the Zhou & Feng paper.
Earlier quoted context omitted.
Not sure where you're getting this from, but CNNs are definitely still state of the art on MNIST. Papers often cite outdated CNN numbers; in fact the best published CNN accuracy numbers are probably lower than they could be - vanilla MNIST for supervised learning is a pretty useless benchmark for computer vision researchers now.
My bad, I should have been more specific. The Deep PCA paper shows superior performance not on baseline MNIST, but on the MNIST variations. I also agree that vanilla MNIST is pretty useless for Computer Vision researches and was trying (awkwardly) to support that idea by showing how these other non Deep Learning techniques performed equally well.
So using 80 times more machines makes you 60 times faster (assuming those are the same machines) "while performing better on 23 games tested, and worse on 28"[0]?
[0] The paper for this blog post https://arxiv.org/pdf/1703.03864.pdf
Asynchronous advantage actor critic (A3C) https://arxiv.org/pdf/1602.01783.pdf