fynance.models.rolling.RollMultiLayerPerceptron

class fynance.models.rolling.RollMultiLayerPerceptron(X, y, layers=[], activation=None, drop=None, bias=True, x_type=None, y_type=None, activation_kwargs={}, **kwargs)

Rolling version of the vanilla neural network model.

Attributes:

lr_scheduler

optimizer

seed_numpy

seed_torch

Methods

run([backtest_plot, backtest_kpi, figsize, func])	Run neural network model and backtest predictions.
set_roll_period(train_period, test_period[, …])	Callable method to set target features data, and model.
sub_predict(X)	Predict.
save(path)	Save the trained neural network model.

TODO:
- fix train and predict methods
- finish docstring
- finish methods