Comparison between likelihood and average of sum of outcomes based reward functions (research objective) (Using deep reinforcement learning for personalizing review sessions on e-learning platforms with spaced repetition)
Here the authors aimed to measure the difference between likelihood and average of sum of outcomes based reward functions. The intuition behind using average of sum of outcomes was that it would it would release students states from the reward functions dependency. Average of sum of outcomes had relatively similar performance with the likelihood reward functions when used with EFC and HLR, but the same can't be said about GPL as the performance fluctuated. As it was described the reward distributions of these functions were quite different. For p < .001 for GPL student model average of sum of outcomes had slightly better result than the likelihood one.
0
1
Tags
Data Science
Related
Relation between rewards and thresholds (Using deep reinforcement learning for personalizing review sessions on e-learning platforms with spaced repetition)
Performance of DRL agent when the number of items are varied (Using deep reinforcement learning for personalizing review sessions on e-learning platforms with spaced repetition)
Comparison of Performance of TRPO and TNPG algorithms (Using deep reinforcement learning for personalizing review sessions on e-learning platforms with spaced repetition)
Comparison between likelihood and average of sum of outcomes based reward functions (research objective) (Using deep reinforcement learning for personalizing review sessions on e-learning platforms with spaced repetition)
Performance of TRPO with reward shaping (research objective) (Using deep reinforcement learning for personalizing review sessions on e-learning platforms with spaced repetition)