eb4ffb6cdaf6f04f978fc57e32c95c8b4a33bcb6,examples/federated_learning_with_encryption.py,Client,fit,#Client#,117

Before Change

        //
        //     print("Error %.4f" % mean_square_error(self.predict(X), y))

        self.weights = np.linalg.inv(X.T.dot(X)).dot(X.T).dot(y)
        // print(self.weights)

        return self

After Change

        for _ in range(n_iter):
            for i in range(length):
                delta = self.predict(X[i, :]) - y[i]
                for j in range(dim):
                    self.weights[j] -= eta * delta * X[i, j]

            // print("Error %.4f" % mean_square_error(self.predict(X), y))

        // self.weights = np.linalg.inv(X.T.dot(X)).dot(X.T).dot(y)
        // print(self.weights)

        return self

    // def compute_gradient(self, X, y):
    //     Return the gradient computed at the current model.

Instances