| Topic: | [Deep Learning] Bayesian Learning for Binary Neural Networks |
|---|
| Department: | Katedra kybernetiky |
|---|
| Supervisor: | Mgr. Oleksandr Shekhovtsov, Ph.D. |
|---|
| Announce as: | Diplomová práce, Bakalářská práce, Semestrální projekt |
|---|
| Description: | Bayesian learnings paradigm averages over all models that explain the data well. The task will be to compare several methods first on simulated data, where the optimal classification and Bayesian predictive probability can be computed and visualized. Then extrapolate the intuition about the methods to large-scale real problem setting. Binary networks are of special interest as they can be extremely efficient and fast at test time, even when averaging over multiple samples in the Bayesian setting. |
|---|
