SGD Classifier

Model Training

Train the model on the training data

from sklearn.linear_model import SGDClassifier as clf

sgd_model = clf()
sgd_model.fit(X_train, y_train)

Predicting output

The data must be an array of array as that is what the model is trained on.

sgd_model.predict([[<data_array>]])

We need to check the accuracy of the predicted data, which will be taken from the validation data

sgd_model.score(X_valid, y_valid)

Depending on the type of problem, the accuracy can vary to certain degrees, going as low as \(40\%\) or as high as \(90\%\)

The learning is based off of the weights of each data column. In layman terms, the input data of each column in multiplied by the weights of the column to get the prediction. For checking the weights of each input column

sgd_model.coef_