Recipe


Recipe

You can use Recipe to choose some presets for the TimeSequencePredictor fitting process by passing to the recipe field in the fit method.

SmokeRecipe

A very simple Recipe for smoke test that runs one epoch and one iteration with only 1 random sample.

SmokeRecipe()

LSTMRandomGridRecipe

A recipe mixing random and grid using LSTM Model only

LSTMGridRandomRecipe(num_rand_samples=1,epochs=5,training_iteration=10,look_back=2,lstm_1_units=[16, 32, 64, 128],lstm_2_units=[16, 32, 64], batch_size=[32, 64])

Arguments

MTNetRandomGridRecipe

A recipe mixing random and grid using MTNet Model only

MTNetGridRandomRecipe(num_rand_samples=1,
                 epochs=5,
                 training_iteration=10,
                 time_step=[3, 4],
                 long_num=[3, 4],
                 cnn_height=[2, 3],
                 cnn_hid_size=[32, 50, 100],
                 ar_size=[2, 3],
                 batch_size=[32, 64])

Arguments

RandomRecipe

Pure random sample Recipe. Often used as baseline.

RandomRecipe(num_rand_samples=1, look_back=2)

Arguments

GridRandomRecipe

A recipe involves both grid search and random search. The arguments are the same with RandomRecipe.

GridRandomRecipe(num_rand_samples=1, look_back=2)

BayesRecipe

A recipe to search with Bayes Optimization. You need to pre-install bayesian-optimization before using the recipe.

BayesRecipe(num_samples=1, look_back=2)