- 19 Dec, 2019 19 commits
-
-
Vivien Kraus authored
-
Vivien Kraus authored
-
Vivien Kraus authored
-
Vivien Kraus authored
-
Vivien Kraus authored
-
Vivien Kraus authored
This mirror should be less prone to downtime since I am the only one cloning it for now.
-
Vivien Kraus authored
Gnulib is a /huge/ project. Cloning it with each CI is a waste of resources.
-
Vivien Kraus authored
-
Vivien Kraus authored
-
Vivien Kraus authored
-
Vivien Kraus authored
-
Vivien Kraus authored
-
Vivien Kraus authored
If we now bootstrap the perceptron on solo-generated games, we now have an exploration parameter.
-
Vivien Kraus authored
This is a combination of the random AI and a perceptron. This will let us tune the solo difficulty.
-
Vivien Kraus authored
An AI is used to play for the other players, and at the end of the game it learns the game of the main player.
-
Vivien Kraus authored
The next step is to give the solo its own AI implementation.
-
Vivien Kraus authored
-
Vivien Kraus authored
It is not fun when we upgrade the build packages and now all links are somehow broken.
-
Vivien Kraus authored
The =TarotAi= module now factors in all the boring stuff around applying the AI, so we can now remove the code to compute the strongest player in the features module, and integrate the stacking predictor and the perceptron predictor directly in this framework, yet again removing duplicated code to compute the best event (for instance).
-
- 11 Dec, 2019 6 commits
-
-
Vivien Kraus authored
-
Vivien Kraus authored
-
Vivien Kraus authored
-
Vivien Kraus authored
-
Vivien Kraus authored
There is no appstreamcli in guix.
-
Vivien Kraus authored
-
- 08 Dec, 2019 2 commits
-
-
Vivien Kraus authored
-
Vivien Kraus authored
-
- 07 Dec, 2019 13 commits
-
-
Vivien Kraus authored
-
Vivien Kraus authored
-
Vivien Kraus authored
-
Vivien Kraus authored
-
Vivien Kraus authored
-
Vivien Kraus authored
-
Vivien Kraus authored
-
Vivien Kraus authored
-
Vivien Kraus authored
This is superfluous.
-
Vivien Kraus authored
This lets me use commas as separators.
-
Vivien Kraus authored
-
Vivien Kraus authored
-
Vivien Kraus authored
I often get times where all the predictions are set to 0. This is because in the beginning where there is a lot of exploration, the regression is simply based on the bid and whether we are the taker. So, most parts of the network are "discarded", which is to say the weights optimize so that the hidden neuron is never activated. However, at the end of the exploration, we need to exploit these neurons.
-