Experiment (Deep Attentive Study Session Dropout Prediction in Mobile Learning Environment)
- Training Details
The model that gave best AUC score was chosen for the testing phase. The best model had 4 stacked encoder/decoder layers, the latent dimension was 512, the number of heads in multi head attention layer was equal to 8. In this model Xavier initialization was utilized and Adam optimizer was used.
- Experiment Results
Besides DAS, LSTM and GRU networks were used and they get as an input and . Different sequence size was tried for all of this models: 5 and 25.
0
1
Tags
Data Science
Related
Introduction (Deep Attentive Study Session Dropout Prediction in Mobile Learning Environment)
Related Work (Deep Attentive Study Session Dropout Prediction in Mobile Learning Environment)
Study Session Dropout in Mobile Learning (Deep Attentive Study Session Dropout Prediction in Mobile Learning Environment)
Propose Method (Deep Attentive Study Session Dropout Prediction in Mobile Learning Environment)
Dataset (Deep Attentive Study Session Dropout Prediction in Mobile Learning Environment)
Experiment (Deep Attentive Study Session Dropout Prediction in Mobile Learning Environment)