Important papers and blogs on Generative modelling

classical generative model :

1. https://cs229.stanford.edu/notes2021fall/cs229-notes8.pdf (Note on EM Algoithm,ELBO maximisation from stanford)
   
2. https://youtu.be/LmpkKwsyQj4 (Anant Avati's lecture on K-means--> GMM-->>EM algorithm,A Must watch)
   
3. https://cs229.stanford.edu/summer2019/cs229-notes9.pdf (Factor analysis (A latent space model) lecture notes).
   
4. https://www.countbayesie.com/blog/2017/5/9/kullback-leibler-divergence-explained(A good blog that explains about KL Divergence)
   
5. https://stillbreeze.github.io/Variational-Inference-and-Expectation-Maximization/ (On Expectation Maximization)

Vanilla VAE :

1. http://stillbreeze.github.io/Variational-Inference-and- Expectation-Maximization/
2. https://www.cs.cmu.edu/~tom/10-702/Zoubin-702.pdf
3. https://arxiv.org/pdf/1312.6114.pdf (VAE base paper)
4. https://arxiv.org/pdf/1606.05908.pdf (A tutorial on Variational Auto-Encoder)
5. https://gregorygundersen.com/blog/2018/04/29/ reparameterization/(A really good blog on reparametrisation)
6. http://approximateinference.org/accepted/HoffmanJohnson2016.pdf (ELBO Paper)

Some of the important extensions over Vanilla VAE:

1. https://openreview.net/references/pdf?id=Sy2fzU9gl (Beta -VAE base paper)
2. https://arxiv.org/pdf/1804.03599.pdf (discusses about dissentanglement in Beta-VAE)
3. https://arxiv.org/pdf/1706.02262.pdf (Info- VAE)
4. https://arxiv.org/abs/1705.07120 (VAE with vamp Prior)
5. https://ml.berkeley.edu/blog/posts/vq-vae/ (Vector quantised VAE)

Matrix completion approach(Classical): SVD,CUR,PQ

1. Good paper on randomised SVD (recent) : https://arxiv.org/pdf/1608.02148.pdf
2. CUR decomposition Base paper : https://www.pnas.org/doi/full/10.1073/pnas.0803205106
   
3. Good material for understanding CUR better (recent) : https://arxiv.org/abs/1907.12668
   
4. PQ Decomposition blog : https://towardsdatascience.com/recommendation-system-matrix-factorization-d61978660b4b
   
5. Stanford mini lecture(inside a playlist) : https://youtu.be/SO1KTzuKTSI
   

Neural Collaborative Filtering(NCF,Deep learning based):

1. NCF base paper : https://arxiv.org/pdf/1708.05031.pdf

Some blogs that I followed (Will make life easier)

• https://jxmo.io/posts/nce (On NCE)
  
• https://github.com/eugeneyan ()
  
• https://eugeneyan.com/writing/bandits/
  
• https://uvadlc-notebooks.readthedocs.io/en/latest/tutorial_notebooks/tutorial8/Deep_Energy_Models.html
• https://www.robots.ox.ac.uk/~ojw/files/NotesOnCD.pdf (On Contrastive Divergence)
• https://timvieira.github.io/blog/post/2014/12/21/importance-sampling/ (On importance sampling)
• https://timvieira.github.io/blog/post/2016/12/19/counterfactual-reasoning-and-learning-from-logged-data/ (off-policy)
• https://web.archive.org/web/20170401030417/http://www.columbia.edu/~mh2078/MCS04/MCS_var_red2.pdf(Variance reduction techniques)
• https://jonathan-hui.medium.com/gan-why-it-is-so-hard-to-train-generative-advisory-networks-819a86b3750b (Really good blog that portraits the usefulness of JS divergence(Vanilla GAN) instead of KL divergence(VAE) )

Important papers on GAN

• Generative Adversarial Networks(https://arxiv.org/abs/1406.2661)
• Adversarial Feature Learning(https://arxiv.org/abs/1605.09782)
• DC-GAN (https://arxiv.org/abs/1511.06434)
• GAN vs Actor Critic (https://www.quora.com/How-is-the-GAN-different-from-the-actor-critic-method-in-reinforcement-learning)
• Modifying naive GAN training(Alternative GD) procedure.Many papers got published in ICLR-2016,2017,2018 proposing various other training procedure.This is the paper that introduces issues like modal collapse (https://arxiv.org/pdf/1611.02163.pdf)
• Chekhov GAN-offeres a game theortic inspired approach for training the GAN. (https://arxiv.org/pdf/1706.03269.pdf?).
• Link to a preesentation slides made by me on the chekhov GAN.(https://docs.google.com/presentation/d/1NdsimG3mRsrfH6hH3O4PYxbOllFjxRlobkD09a8GoJU/edit?usp=sharing)

Score-Based models:

Will be updated soon ...

Energy Based models:

Will be updated soon ...

Diffusion models:

• Understanding Diffusion Models: A Unified Perspective:An important paper that connects VAE to diffusion through Heirarchical VAE . (https://arxiv.org/abs/2208.11970)

RL-reference materials:

• https://huggingface.co/blog/deep-rl-pg (Deep+RL)
• https://jonathan-hui.medium.com/rl-policy-gradients-explained-9b13b688b146 (Policy gradient-1,2)
• https://www.davidsilver.uk/teaching/ (David silver-lectures)
• https://sites.google.com/view/deep-rl-bootcamp/lectures (Peter Abbeil lectures)