Deep Learning
Posts and notes about deep learning.
Series & Posts
1
Neural networks explained: neurons, activations, and layers
2 Backpropagation: forward pass, chain rule, and gradient flow
3 Training neural networks: initialization, learning rates, and schedules
4 CNNs explained: convolution, pooling, and feature maps
5 RNNs and LSTMs: sequence modeling and gating mechanisms
6 Attention and transformers: self-attention, multi-head, and positional encoding
7 Word embeddings: Word2Vec, GloVe, and FastText explained
8 Transfer learning and fine-tuning: from ImageNet to your task
9 DNN optimization: momentum, Adam, and learning rate schedules
10 Deep learning regularization: dropout, batch norm, and data augmentation
11 Encoder-decoder architecture: seq2seq and U-Net explained
12 Generative models overview: autoregressive, flow-based, and latent variable
13 RBMs explained: energy-based models and contrastive divergence
14 Deep Belief Networks: greedy layer-wise pretraining explained
15 VAE explained: ELBO, reparameterization trick, and KL divergence
16 GANs explained: minimax game, mode collapse, and Wasserstein distance
17 DCGAN, conditional GANs, CycleGAN, and StyleGAN explained
18 Representation learning: autoencoders, contrastive learning, and SimCLR
19 Domain adaptation: covariate shift, DANN, and few-shot learning
20 Distributed representations: latent spaces and embedding geometry
21 AutoML: grid search, Bayesian optimization, and Hyperband
22 Neural Architecture Search: reinforcement learning, evolution, and DARTS
23 Model compression: pruning, quantization, and knowledge distillation
24 Graph Neural Networks: GCN, GraphSAGE, and GAT explained
25 Debugging deep learning: loss curves, gradient health, and ablations
26 ResNet architecture explained: skip connections and residual learning