Dataset: CIFAR 10 Due: Dec 13
Report: 4 pages long
Elements:
- Data exploration.
- Two ML techniques.
- The design choices should be based on what we learned by exploring the data.
- We should also mention the reasons why we chose those hyperparameters, e.g. to control complexity.
- Feature design or model optimization.
- Data augmentation?
- Performance validation.
- Cross validation, model selection, etc.
- Summary of how we changed the models to prevent under- or over-fitting.