Why is a single Perceptron unable to solve the XOR gate problem?

Because the XOR function is not linearly separable. A single straight decision boundary cannot separate (0,1) and (1,0) from (0,0) and (1,1). Hidden layers are required.

What is the function of the bias term in an artificial neuron?

The bias allows the neuron to shift the activation function trigger threshold left or right, independent of the input feature values.

Home ASSEB AI HubUnit 1: Artificial Neural Networks

Class XIIASSEB-527Unit 1: Artificial Neural Networks

Unit 1: Artificial Neural Networks

Master official ASSEB Class XII AI Unit 1: Perceptrons, Multilayer Perceptrons (MLP), Backpropagation, Activation Functions (Sigmoid, ReLU, Softmax), and Scikit-learn ANN housing regression.

Theory: 8 Marks

•

Practical: 5 Marks

•

12 min read

Subscription Gate Active

Learning Outcomes

Understand the structure and biological inspiration of Artificial Neural Networks (ANNs).
Analyze the single Perceptron model, linear decision boundaries, and its limitation with the non-linear XOR problem.
Design Multilayer Perceptrons (MLPs) incorporating hidden layers and non-linear activation functions (Sigmoid, Tanh, ReLU, Softmax).
Master the Backpropagation algorithm: chain rule gradient propagation, loss functions (MSE, Cross-entropy), and iterative weight updates.
Implement an end-to-end ANN regression model in Python using Scikit-learn (MLPRegressor) to predict housing prices.

Board Exam Scoring Tips

In board exams, clearly write the perceptron equation: $y = f(\sum_{i=1}^n w_i x_i + b)$ where $w$ is weight, $x$ is input, and $b$ is bias.
Expect a 3-mark question explaining why Backpropagation is necessary in deep networks (propagates error backward via the chain rule to compute partial derivatives).
Always normalize input features using `StandardScaler` before feeding them into an MLP.

Key Takeaways

The fundamental building block of an ANN is the artificial neuron, which computes a weighted sum of inputs plus bias and applies an activation function.
A single Perceptron acts strictly as a linear classifier; adding hidden layers with non-linear activation functions enables MLPs to solve complex non-linear problems.
The iterative learning cycle of an ANN consists of: 1) Forward pass, 2) Loss calculation, 3) Backpropagation of error gradients, and 4) Weight and bias updates.
ReLU (Rectified Linear Unit) avoids the vanishing gradient saturation that affects Sigmoid in deep hidden layers.
Scikit-learn provides accessible ANN implementations (`MLPClassifier` and `MLPRegressor`) for higher secondary lab practicals.

Practice Unit MCQ Quiz

Back to ASSEB AI Hub

# Python Errata Fix Sample...

Premium ASSEB Syllabus

Unlock Unit 1: Artificial Neural Networks

Access verified code errata manual, 50 external viva questions, period lesson scripts, and practical examination rubrics.

Student Pass₹149/mo

Full Class XIIunits, non-math question bank & Doubt Desk.

Teacher Pass₹299/mo

All XI & XII units, 45-min lesson scripts & 50-mark rubrics.

Single Chapter₹49 once

Instant permanent unlock for this specific chapter and errata code.

Your Email (for watermarking)

WhatsApp Phone

256-Bit Encrypted Payment • Supports UPI, Cards & NetBanking

Previous UnitUnit 8: AI Ethics and Values Next UnitUnit 2: Deep Neural Networks