Neural Networks
Perceptron & Feedforward Neural Networks
Basic building blocks of neural models.
Learn weighted combinations of inputs.
Foundation of all modern neural networks.
Activation Functions
Introduce non-linearity into networks.
Enable learning of complex patterns.
Strongly affect training dynamics.
Backpropagation
Algorithm for computing gradients efficiently.
Enables learning through error correction.
Core mechanism behind neural network training.
Deep Neural Networks
Networks with multiple hidden layers.
Capture hierarchical feature representations.
Power many modern AI applications.
Convolutional Neural Networks (CNNs)
Specialized for grid-like data such as images.
Learn spatial and local patterns efficiently.
Widely used in computer vision.
Recurrent Neural Networks (RNNs) & Transformers
Designed for sequential and temporal data.
Handle language, time series, and signals.
Transformers dominate modern NLP tasks.
Training Techniques & Optimization for NNs
Includes learning rates, batch normalization, and dropout.
Improves convergence and stability.
Essential for scaling deep models.
