Add option to not shuffle rows

[tmontana]

When dealing with time series data it is important to not shuffle rows when performing cross-validation. Would be good to have that as an option.

[pplonski]

Right now, as default the 5-fold cross-validation is used for assesing model's performance. To change this behavior we need to add validation variable to AutoML interface.

[pplonski]

This can be done by setting the _validation variable after AutoML object is constructed. Example:

automl = AutoML()
automl._validation = {"validation_type": "kfold", "k_folds": 5, "shuffle": False}
automl.fit(X, y)