train dsi data with ml emulator #196
Description
- need to train data using an ml emulator
- data must be in a dsi-compatible db and stable
- users should be able to select which table to train on
- users can select whether do classification or regression predictions
- users can select which columns to use as input and which as output
- users can select types of ml models to train on -- start with 4: decision tree, random forest, knn, linear regression
- users can select specific metrics to evaluate -- different if classification or regression
outputs
- output table of metrics across different models with optimal hyperparameters for each selected model on that data
- graphs comparing the metrics of each selected model on that data