HyperParameter Optimization
seqlearner.EmbeddingHyperOptimization.optmize(embedding, sequence_datapath)
Hyper-parameter optimization for an embedding method implementation method.
You can specify the embedding method to function optimize and the best choice for parameters.
The optimize function takes the following arguments:
Arguments
- embedding: String, Embedding method which its hyper-parameters are going to be optimized
- sequence_datapath: String, sequences file path
The optmize function returns a dictionary of hyperparamters and their best value for the corresponding hyperparameters.
Example: Hyperparameter optimization for Freq2Vec
from seqlearner import EmbeddingHyperOptimization as eho
labeled_path = "../data/labeled.csv"
unlabeled_path = "../data/unlabeled.csv"
best_parameters = eho.optimize(embedding="freq2vec")
print(best_parameters)
This will print the dictionary of best values for each hyper-parameter attending in Freq2Vec embedding method.