Applications of Computational Intelligence Methods
This seminar is currently taught in Czech only.
The source codes for the lessons of the seminar are available at GitHub (in Czech only). You can also browse the notebooks with the notes and pictures there.
Tentative schedule
| Date | Topic |
|---|---|
| 6. 10. | Visualization and model evaluation |
| 13. 10. | Feature extraction and selection |
| 20. 10. | Linear models |
| 27. 10. | Kernel methods |
| 3. 11. | Deep networks (convNet, Stacked AE) |
| 10. 11. | Deep networks (LSTM, word2vec) |
| 24. 11. | Deep reinforcement learning |
| 1. 12. | Ensembles |
| 8. 12. | Semi-supervised learning |
| 15. 12. | Meta-learning (grid search, evolution) |
| 5. 1. | Combination of EA and ML (surrogate models) |
| 12. 1. | CMA-ES + constraints in EA |
Credit requirements
You will obtain the credit for team work during the seminar. The idea is that you will work in teams of 2-3 students and each group will have around 4 presentation during the term (approx. 5 minutes each) about the results you were able to obtain using the methods from the seminar. The goal is to show, what you could do and provide inspiration for others hot to solve problems and present the results.