Matching Model and Data
Finding the right balance between model complexity and data
A good AI system needs:
- the right amount of data
- the right model complexity
The model must fit the data, and the data must fit the task.
The model architecture defines the complexity of the model.
If the balance is wrong, the model may face training limits.
Depending on the problem, different:
- models
- loss functions
- training methods are used to reach the best result.