Traffic speed forecasting

The dataset is saved on Python's Numpy library format.The dataset is splitted in training and testing.X is the inputs and Y is the outputs.X is a feature-normalized 3D matrix (samples, time intervals, features).Y is a 2D matrix (samples, time intervals).*X features are:speed_avg_1: Average speed on lane 1speed_avg_2: Average speed on lane 2speed_avg_3: Average speed on lane 3speed_med_1: Median speed on lane 1speed_med_2: Median speed on lane 2speed_med_3: Median speed on lane 3speed_min_1: Minimum speed on lane 1speed_min_2: Minimum speed on lane 2speed_min_3: Minimum speed on lane 3speed_max_1: Maximum speed on lane 1speed_max_2: Maximum speed on lane 2speed_max_3: Maximum speed on lane 3speed_std_1: Speed's standard deviation on lane 1speed_std_2: Speed's standard deviation on lane 2speed_std_3: Speed's standard deviation on lane 3volume_1: Volume on lane 1volume_2: Volume on lane 2volume_3: Volume on lane 3detected_1: (dummy) vehicle detected on lane 1detected_2: (dummy) vehicle detected on lane 2detected_3: (dummy) vehicle detected on lane 3day_week_1: (dummy) day of the week = mondayday_week_2: (dummy) day of the week = tuesdayday_week_3: (dummy) day of the week = wednesdayday_week_4: (dummy) day of the week = thursdayday_week_5: (dummy) day of the week = fridayprecipitation: rain intensity*Y is the average speed of the time interval considering all lanes*days is a vector to indicate a day ID for each sample.*u and s are the average and standard deviation used for normalization of X_train and X_test and calculated based on X_train