Course

Unit I

Biological neurons and brain inspiration. Artificial neuron concept. McCulloch–Pitts neuron model. Threshold logic units. Linear separability.

Unit II

Learning paradigms: Supervised learning, Unsupervised learning. Hebbian learning rule. Perceptron learning algorithm. Adaline and LMS rule. Limitations of single‑layer perceptrons.

Unit III

MultiLayer Neural Networks. Need for hidden layers. Multi‑Layer Perceptron (MLP) architecture. Activation functions: Sigmoid, Tanh, ReLU (conceptual introduction only). Network capacity and representation power.

Unit IV

Training Neural Networks. Error functions. Backpropagation algorithm (conceptual and mathematical overview). Gradient descent learning. Learning rate, momentum. Convergence behavior. Overfitting and generalization.

Unit V

Pattern classification. Function approximation. Associative memory. Hopfield networks (intro). Limitations of neural networks: Local minima. Interpretability. Data dependency. Ethical considerations in neural decision systems.

Course Objective(s) 

• Introduce the biological inspiration and mathematical foundations of neural networks. 
• Develop an understanding of basic neuron models and learning rules. 
• Enable students to analyze single‑layer and multi‑layer neural network architectures. 
• Explain training mechanisms, convergence issues, and limitations of neural networks. 
• Provide a conceptual foundation for advanced neural and deep learning courses 

Course Outcomes 

CO-PO Mapping 

• Neural Networks and Learning Machines, 3rd Edition by Simon Haykin. Pearson, 2009.
• Artificial Neural Networks by B. Yegnanarayana. PHI Learning, 2009.