Scikit Linear Regression Model
This issue discusses the integration of regression model in iota2. It is related to: #79, #325 and #487
Ridge Linear Regression (or Ordinary least square)
Use the model RidgeCV from sklearn.
The parameters alphas
and cv
should be exposed to the user in the configuration file. The former should be list of strictly positive numbers and the latter should be a positive number (smaller than the number of sample in the data-set) or None
(default) value.
Lasso
Use the model LassoCV from sklearn.
The parameters alphas
and cv
should be exposed to the user in the configuration file. The former should be list of strictly positive numbers and the latter should be a positive number (smaller than the number of sample in the data-set) or 5
(default) value.
HuberRegressor
Use the model HuberRegressor from sklearn.
The parameters epsilon
should be exposed to the user in the configuration file. It should be a float number greater than 1.0 and default to 1.35.