- https://github.com/floodsung/Deep-Learning-Papers-Reading-Roadmap - roadmap for NN from basis to advance
- https://paperswithcode.com/ - papers with code
- https://andlukyane.com/blog/ - review of modern papers
- https://farid.one/kaggle-solutions/ - top kaggle solutions
- https://emacsway.github.io/ru/self-learning-for-software-engineer - #offtop computer science roadmap
- http://jalammar.github.io - best visualisations and explanation for papers
- https://lena-voita.github.io/nlp_course.html - excelent NLP course pages
- http://nlpprogress.com/ - SOTA rank results for every NLP task
- Word2Vec, 2013
- Negative Sampling, Hierarchical Softmax in Word2Vec, 2013
- FastText, 2016
- Clear explanation FastText
- BERT, 2018
- SummaRuNNer - text summarization
- P-tuning - train embeddings for prompt tokens and freeze main LLM params to solve specific tasks
- Object Detection in 20 Years: A Survey
- SSD - best tutorial
- The Unreasonable Effectiveness of Recurrent Neural Networks
- Neural Machine Translation by Jointly Learning to Align and Translate - attention in machine translating with RNN
- Deep Reinforcement Learning: Pong from Pixels - policy function algorithm
- *Playing Atari with Deep Reinforcement Learning - Q-function algorithm
- Distilling the Knowledge in a Neural Network
- *A Survey of Quantization Methods for Efficient Neural Network Inference
- *Modeling Teacher-Student Techniques in Deep Neural Networks for Knowledge Distillation
- Empirical_Study_of_Uniform_Architecture_Knowledge_Distillation
- Batchnorm explanation - main idea for cv: norm over minibatch for each filter
- Layernorm explanation - main idea for cv: norm over ONE sample in minibatch for each filter
- Guide for learning rate schedulers
- A Recipe for Training Neural Networks - tricks and guide for coding DL|ML
- Deep Double Descent: Where Bigger Models and More Data Hurt
- Auto-Encoding Variational Bayes (VAE)
- VAE notes - excelent explanation for VAE in russian
- Attention Is All You Need - transformer
- Explanation of "Attention Is All You Need" - best explanation of paper
- *GPT-3
- Collaborative Filtering for Implicit Feedback Datasets - SGD, MSE, implicit, clean dataset, explain recommendations
- ALS distributed, notes
- Bayesian Personalized Ranking from Implicit Feedback - LearnBPR, MF, bayes optimizing ranking function
- WSABIE: Scaling Up To Large Vocabulary Image Annotation - smart pairwise sampling, optimizing ranking function
- *Factorization Machine
- *Improving recommendation lists through topic diversification
- *Who likes it more Mining Worth-Recommending Items from Long Tails by Modeling Relative Preference
- Monolith: Real Time Recommendation System With Collisionless Embedding Table - tik-tok online learning, cucko hash encoding
- Deep Neural Networks for YouTube Recommendations
- BERT4Rec
- SAS4Rec
- Trustworthy Online Controlled Experiments - A Practical Guide to A/B Testing
- Practitioner’s Guide to Statistical Tests - tests for proportion, pitfalls
- Possion bootstrap
- *Delta Method
- Hypo testing bootstrap
- CUPED at A/B. Paper
- CUPED, X5 simulations
- Linearization at A/B
- Overlapping Experiment Infrastructure: More, Better, Faster Experimentation - several experiments for one user at once
- Identifying Ambiguous Queries in Web Search
- Determining the user intent of web search engine queries
- Embedding-based Retrieval in Facebook Search - embedding search
- Que2Search - improving Embedding-based Retrieval in Facebook Search with bert model
- A Survey of Model Architectures in Information Retrieval
- SPLADE - good vizualitaion
- Real-time Personalization using Embeddings for Search Ranking at Airbnb (Airbnb 2018)
- Personalized Transformer-based Ranking for e-Commerce at Yandex
- Real-Time Personalized Ranking in E-commerce Search - 3 types of features, Kendall-Tau measuring personalization
- Can clicks be both labels and features? Unbiased Behavior Feature Collection and Uncertainty-aware Learning to Rank
- When Inverse Propensity Scoring does not Work - Affine Corrections for Unbiased Learning to Rank - first introduction affince corrections
- Addressing Trust Bias for Unbiased Learning-to-Rank - EM algorithm explanation
- Position Bias Estimation for Unbiased Learning to Rank in Personal Search - IPS
- Don’t Forget This: Augmenting Results with Event-Aware Search
- Playlist Search Reinvented: LLMs Behind the Curtain
- Ranking Across Different Content Types: The Robust Beauty of Multinomial Blending
- Extremely efficient online query encoding for dense retrieval
- *Causal Inference and Uplift Modeling. A review of the literature
- Adapting Neural Networks for Uplift Models
- Practical Lessons from Predicting Clicks on Ads at Facebook - downsampling, features from GBM for LogReg, data freshness
- Smart Pacing for Effective Online Ad Campaign Optimization
- F1-score, ROC_AUC, PR-AUC, comparison, prons and cons
- NDCG
- MRR - Mean_reciprocal_rank
- Kats - time series library (better than prophet)