Tips - Chosen computational solution

[rifqialf]

The group added basic statistics of each bands as features and also used median for the indices. More reasoning for these decisions might add more context to the machine learning process.

[rifqialf]

The non-dask implementation used Random Forest while the dask implementation used Multi-layer Perceptron. This difference makes it difficult to observe how dask influence the machine learning process. It might be better to stick to one algorithm to make both implementations comparable.

[rifqialf]

Selecting Random Forest and Multi-layer Perceptron for this task is interesting. More assignment-specific reasoning might be helpful to add more contect to the added general justification of these ML algorithm choices.