Lab 11: Neural Network (Backpropagation) for Classification

Aim

Train a simple MLP on XOR to observe non-linear decision boundaries.

Objectives

• Define XOR dataset.
• Train MLPClassifier with one hidden layer.
• Check predictions on training set.

Algorithm / Procedure

1. Create inputs and labels for XOR.
2. Instantiate MLPClassifier(hidden_layer_sizes=(5,), max_iter=1000).
3. Fit and print predictions.

Python Code

from sklearn.neural_network import MLPClassifier

X = [[0,0],[0,1],[1,0],[1,1]]
y = [0,1,1,0]

nn = MLPClassifier(hidden_layer_sizes=(5,), max_iter=1000, random_state=0).fit(X, y)
print("Predictions:", nn.predict(X).tolist())
print("Training accuracy:", nn.score(X, y))

Sample Output (expected)

Predictions: [0, 1, 1, 0]
Training accuracy: 1.0