- ee559-handout-1-1-from-anns-to-deep-learning.pdf
- ee559-handout-1-2-current-success.pdf
- ee559-handout-1-3-what-is-happening.pdf
- ee559-handout-1-4-tensors-and-linear-regression.pdf
- ee559-handout-1-5-high-dimension-tensors.pdf
- ee559-handout-1-6-tensor-internals.pdf
- ee559-handout-10-1-autoregression.pdf
- ee559-handout-10-2-causal-convolutions.pdf
- ee559-handout-10-3-NVP.pdf
- ee559-handout-2-1-loss-and-risk.pdf
- ee559-handout-2-2-overfitting.pdf
- ee559-handout-2-3-bias-variance-dilemma.pdf
- ee559-handout-2-4-evaluation-protocols.pdf
- ee559-handout-2-5-basic-embeddings.pdf
- ee559-handout-3-1-perceptron.pdf
- ee559-handout-3-2-LDA.pdf
- ee559-handout-3-3-features.pdf
- ee559-handout-3-4-MLP.pdf
- ee559-handout-3-5-gradient-descent.pdf
- ee559-handout-3-6-backprop.pdf
- ee559-handout-4-1-DAG-networks.pdf
- ee559-handout-4-2-autograd.pdf
- ee559-handout-4-3-modules-and-batch-processing.pdf
- ee559-handout-4-4-convolutions.pdf
- ee559-handout-4-5-pooling.pdf
- ee559-handout-4-6-writing-a-module.pdf
- ee559-handout-5-1-cross-entropy-loss.pdf
- ee559-handout-5-2-SGD.pdf
- ee559-handout-5-3-optim.pdf
- ee559-handout-5-4-l2-l1-penalties.pdf
- ee559-handout-5-5-initialization.pdf
- ee559-handout-5-6-architecture-and-training.pdf
- ee559-handout-5-7-writing-an-autograd-function.pdf
- ee559-handout-6-1-benefits-of-depth.pdf
- ee559-handout-6-2-rectifiers.pdf
- ee559-handout-6-3-dropout.pdf
- ee559-handout-6-4-batch-normalization.pdf
- ee559-handout-6-5-residual-networks.pdf
- ee559-handout-6-6-using-GPUs.pdf
- ee559-handout-7-1-transposed-convolutions.pdf
- ee559-handout-7-2-autoencoders.pdf
- ee559-handout-7-3-denoising-autoencoders.pdf
- ee559-handout-7-4-VAE.pdf
- ee559-handout-8-1-CV-tasks.pdf
- ee559-handout-8-2-image-classification.pdf
- ee559-handout-8-3-object-detection.pdf
- ee559-handout-8-4-segmentation.pdf
- ee559-handout-8-5-dataloader-and-surgery.pdf
- ee559-handout-9-1-looking-at-parameters.pdf
- ee559-handout-9-2-looking-at-activations.pdf
- ee559-handout-9-3-visualizing-in-input.pdf
- ee559-handout-9-4-optimizing-inputs.pdf
讲义第二部分:
- ee559-handout-11-1-GAN.pdf
- ee559-handout-11-2-Wasserstein-GAN.pdf
- ee559-handout-11-3-conditional-GAN.pdf
- ee559-handout-11-4-persistence.pdf
- ee559-handout-12-1-RNN-basics.pdf
- ee559-handout-12-2-LSTM-and-GRU.pdf
- ee559-handout-12-3-word-embeddings-and-translation.pdf
- ee559-handout-13-1-attention-memory-translation.pdf
- ee559-handout-13-2-attention-mechanisms.pdf
- ee559-handout-13-3-transformers.pdf
配套的幻灯片下载地址:https://bbs.pinggu.org/thread-9569971-1-1.html
1 – Introduction
2 – Machine learning fundamentals
3 – Multi-layer perceptron and back-propagation
4 – Graphs of operators, autograd, and convolutional layers
5 – Initialization and optimization
6 – Going deeper
7 – Autoencoders
8 – Computer vision
9 – Under the hood
10 – Generative models
11 – Generative adversarial models
12 – Recurrent models and NLP
13 – Attention models