Category Archives: Thesis Progress

One shot environments

I currently writing my “one-shot environments” section of the thesis and rerun old experiments. (Maybe a bug causes this but) my “Deletion” agent performs worst than the Vanilla Agent in the Sinus environments. The problem is that the “Deletion” agent … Continue reading

Posted in Thesis Progress | Leave a comment

Selection strategies (3) with noise

I re-ran all experiments again, but with a certain noise level to the action. The noise has been added to all dimensions of the action with a new random number. The random number comes from a standard deviation with sigma=1 … Continue reading

Posted in Thesis Progress | Leave a comment

split probabilities (Sixhumpcamelback)

recall significance splits samples reward 0.1 167.758 1000.0 0.8391405906445085 0.05 127.989 1000.0 0.5690483573253337 0.001 154.865 1000.0 1.0206215175910123 0.0001 134.718 1000.0 1.0236344912295705 DELETION significance splits samples reward 0.1 61.82 1000.0 0.7980104591540007 0.05 61.894 1000.0 0.9266597255352897 0.01 60.142 1000.0 1.0055977331055612 0.001 49.74 … Continue reading

Posted in Thesis Progress | Leave a comment

split probabilities

I played around with the significance levels of TG. HINT: Significance levels I test the Sinusfunction with the same parameter set than the last experiment for significance = 0.05, 0.01, 0.001, 0.0001 It is good to see that the significance … Continue reading

Posted in Thesis Progress | 1 Comment

Selection strategies in TLS (learning curves)

I created also the learning curves for the results from yesterday which are very interesting. I will discuss that tomorrow with Kurt. For today I will try to create some results of splitting probability according to the significance level of … Continue reading

Posted in Thesis Progress | 3 Comments

Selection strategies in TLS

As I mentioned before I recognized a difference when I changed the strategy of selection. First of all I start debugging my code and this was a hard thing to do but after 2 days of fixing I guess the … Continue reading

Posted in Thesis Progress | Leave a comment

split probability

Michael and Kurt asked my for graph that shows the probability of a split at the “i”th sample I ran 10000 experiments for 1000 samples on the sinus and sixhumcamelback function. The first 100 samples shows significant point, after that … Continue reading

Posted in Thesis Progress | Leave a comment