Neural Networks A Classroom Approach By Satish Kumar.pdf ^new^ -

The Story of AlphaGo

Chapter 7: Convolutional Neural Networks

• Key Concepts: Local receptive fields, weight sharing, stride, padding.
• Mathematics: Convolution as a matrix multiplication (im2col trick).
• Lab: Build a LeNet‑5 style CNN from scratch (no high‑level Keras API) to classify Fashion‑MNIST.

The classroom was filled with a mix of curious and skeptical students. Some had heard of neural networks, while others had not. Professor Kumar started by explaining that neural networks were inspired by the human brain's remarkable ability to learn and adapt. Neural Networks A Classroom Approach By Satish Kumar.pdf

4.2 Evaluation Metrics

• Classification: accuracy, precision, recall, F1, ROC-AUC.
• Regression: RMSE, MAE, R^2.
• Sequence: BLEU, ROUGE, METEOR for translation/summarization.
• Calibration: reliability diagrams, expected calibration error.