ML
Mei Lin
3 weeks ago
Dense but never confusing. Every chapter builds cleanly on the last — you can feel the care in the structure.
Free preview available. Sign in and subscribe to unlock the full book.
Vizuara AI Labs · beginner
Deep Learning Fundamentals
Build a neural network from scratch in pure NumPy.
Master deep learning from the ground up by coding an entire neural network by hand — neurons, forward pass, backpropagation, optimizers, and regularization. Every equation is derived and every line is written in plain NumPy before you ever touch a framework.
38 capsules137 figures~7 hoursby Dr. Raj Dandekar
Read on your Kindle or e-reader
Download the EPUB and read offline — perfect for the train. Works on Kindle, Kobo, Apple Books & more.
00Foundations3 capsules
Foundations — 3 chapters.
01What is deep learning, really?conceptfree10 min02The neuron: a tiny math machineintuitionfree11 min03NumPy and the dot product primercode🔒10 min01Neural Network Architecture3 capsules
Neural Network Architecture — 3 chapters.
04Coding neurons and layerscode🔒11 min05The beauty of NumPy and the dot productcode🔒12 min06Stacking multiple layers togethercode🔒10 min02The Forward Pass5 capsules
The Forward Pass — 5 chapters.
07Implementing the dense layer classcode🔒11 min08Broadcasting and array summationcode🔒12 min09Coding activation functionscode🔒12 min10One full forward pass (no loss yet)code🔒12 min11Coding the cross-entropy lossmath🔒10 min03Optimization and Calculus3 capsules
Optimization and Calculus — 3 chapters.
12Introduction to optimizationconcept🔒11 min13Partial derivatives and gradientsmath🔒11 min14The chain rule: backbone of neural networksmath🔒11 min04The Backward Pass (Backpropagation)9 capsules
The Backward Pass (Backpropagation) — 9 chapters.
15Backpropagation on a single neuronmath🔒13 min16Backpropagation through a layer of neuronsmath🔒11 min17The role of matrices in backpropagationmath🔒11 min18Input derivatives in backpropagationmath🔒12 min19Coding the backpropagation building blockscode🔒11 min20Backpropagation through ReLUcode🔒11 min21Backpropagation through cross-entropy lossmath🔒13 min22Combined softmax + cross-entropy backwardmath🔒12 min23Building the full backpropagation pipelinecode🔒11 min05The Full Training Loop1 capsules
The Full Training Loop — 1 chapter.
24The entire forward-backward pass in Pythonproject🔒13 min06Optimizers6 capsules
Optimizers — 6 chapters.
25Coding the gradient descent optimizercode🔒13 min26Learning rate decayconcept🔒11 min27Momentum in trainingintuition🔒11 min28Coding the AdaGrad optimizercode🔒10 min29Coding the RMSProp optimizercode🔒12 min30Coding the Adam optimizercode🔒14 min07Recap: A Network from Scratch1 capsules
Recap: A Network from Scratch — 1 chapter.
31Neural networks from scratch, end to endconcept🔒11 min08Regularization and Testing4 capsules
Regularization and Testing — 4 chapters.
32Testing, generalization, and overfittingconcept🔒11 min33K-fold cross validationconcept🔒11 min34L1 and L2 regularizationmath🔒11 min35Dropout layerscode🔒12 min09Hands-On Projects3 capsules
Hands-On Projects — 3 chapters.
36Regression project: California Housingproject🔒12 min37Classification project: Fashion-MNISTproject🔒10 min38What we built, and where to go nextconcept🔒10 minRatings & reviews
740 readersRate this book
FZ
Fatima Zahra
4 months ago
Practical and rigorous at the same time. I went straight from reading a capsule to shipping it at work.
SR
Sofia Reyes
7 months ago
The hand-drawn diagrams make abstract ideas click instantly. Wish every technical book was written like this.