motus/SCRIPTS_RULES.txt
733246f55d40ml_project2

SCRIPTS_RULES.txt
View Options

			SCRIPTS RULES
			- Metrics used to evaluate each run are: mse and R^2
			- Splitting ratios: training = 0.6, testing = 0.2, validation = 0.2 (referring to total database)
			- Ridge regression deg polynomial expansion = 3
			- Y vector only u_x and u_y (predicting only horizontal velocity)
			- Discretization of y output: KbinDiscretizer with 'uniform' strategy
			- Random State of split_test_train is 50
			- Splitting the data first splitting with train_test_split into training and 'temp' dataset; then consequently split temp into test and evaluation set. (first test_size=0.4, second test_size=0.5).
			- Ridge alphas optimizer = np.logspace(-10,5,200)