further model testing #8
Conversation
|
What are the differences between the two metrics, MAE and RMSE? How do they relate to precision and accuracy? How can we explain this in layman's terms, apply it directly to Helios, and make real world decisions from this? |
promatty
force-pushed
the
ML-11-Experiment-with-various-machine-learning-algorithms
branch
from
74327db to
e02a582
Compare
analysis/furtherModelTesting.ipynb
Outdated
| "\n", | ||
| "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=global_test_size, random_state=global_random_state)\n", | ||
| "# Define the parameter grid for Random Forest\n", | ||
| "param_grid_rf = {\n", |
There was a problem hiding this comment.
Tell me how many times this model will be trained with this size of param_grid
There was a problem hiding this comment.
How would this change if we also wanted to use Cross validation of say... 5?
There was a problem hiding this comment.
the model is trained once for each combination, so it's 3x4x3x3=108 times. cross validation of 5 means we would train 5 times per combination so 5x108= 540. Is this too much?
|
Also I was hoping you would do the hyper parameter tuning optimizations in I feel more comfortable reviewing smaller PRs |
promatty
force-pushed
the
ML-11-Experiment-with-various-machine-learning-algorithms
branch
from
1a04102 to
499c367
Compare
promatty
force-pushed
the
ML-11-Experiment-with-various-machine-learning-algorithms
branch
from
499c367 to
697deae
Compare
justin-phxm
deleted the
ML-11-Experiment-with-various-machine-learning-algorithms
branch
trained models on KNN, Decision Trees, Random Forests, and Gradient Boosting, all with and without Cross Validation.