Multiclass Logistic Regression (aolearn.multiclass_logistic)

fit_multiclass_logistic(X, y, numClasses, learningRate, numIterations)

Trains a multiclass logistic regression model (utilizing softmax).

Parameters:

• X (table): Training features.
• y (table): Training labels (integers representing the different classes).
• numClasses (number): The total number of classes.
• learningRate (number): Learning rate.
• numIterations (number): Number of iterations.

Returns:

• table, table: Learned weights (a table of tables, one for each class) and biases (a table, one for each class).

---

predict_multiclass_logistic(weights, bias, features)

Predicts probabilities for each class.

Parameters:

• weights (table): Learned weights (a table of tables) from fit_multiclass_logistic.
• bias (table): Learned biases from fit_multiclass_logistic.
• features (table): Features for a single data point.

Returns:

• table: Probabilities for each class (table indices correspond to class labels).