Machine Learning Guide
30 Episodes of Machine Learning content. Best audio resource I've found.
How to deliver on Machine Learning projects
Step-by-step pipeline from idea to production for an ML project
Analyzing 50k fonts using deep neural networks
Perfect dataset for training fonts, trained model even has 40 latent factors.
Artificial Intelligence: The Revolution Hasn't Happened Yet
Human-imitative AI is not a good way to frame current progress in AI. Intelligence Augmentation and Intelligence Infrastructure present an equally important (and as-yet-unnamed) class of problems.
#47 - Catherine Olsson & Daniel Ziegler on the fast path into high-impact ML roles
OpenAI and Google Brain engineers on their unconventional paths into AI and the impactful work they're doing
Local Governments Power Up to Advance China's National AI Agenda
China local governments propose aggressive plans totaling $400bn in AI money by 2030!
Documents OCR: Improving Efficiency by Making PDFs Searchable
Use Google Cloud Vision. Good OCR pipeline reference for startup.
Variational Autoencoders Explained
Variational autoencoders are generative encoder-decoder networks with a constraint on the encoding network. Well-explained.
Approaching (Almost) Any Machine Learning Problem
Mental model for approaching ML problems. Very good breakdown.
Ash Fontana -- Investing in Artificial Intelligence
A top AI investor gives his perspective on the industry and his investment thesis
AI Index Thoughts
Thoughts on the AI Index
AI Index Index
Summarizing a summary of the state of AI
From Research To Practice
Notes on Best Practices for Applying Deep Learning to Novel Applications
Language Modeling Survey
Notes on Exploring the Limits of Language Modeling
Who is MiningLamp? Why was it able to win Tencent's high-value investment?
MiningLamp has gotten significant attention and investment for its police-assisting AI.
Winograd Schema Challenge
The Winograd Schema Challenge aims to address the flaws in the Turing Test to determine if an AI is "human level".
GRUs vs. LSTMs
Notes on Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling
#102 - Andrew Kortina
Founder of Venmo and Fin. Future of work.
Markov Decision Processes
Markov Decision Processes are Finite State Machines with four key components: state, action, transition function, reward function. They run into the curse of dimensionality.
Notes on Bag of Tricks for Efficient Text Classification
Transfer learning is powerful and underutilized.
From Research to Startup, There and Back Again
Decades of experience in Silicon Valley with John Hennessy (current chairman of Alphabet)
Be more rigorous in real life by thinking in terms of priors, posteriors, and updates.
Google Rules of Machine Learning
Bite-size, Google-scale advice for ML.
Game theory is the study of equilibria-based solutions.
Reinforcement learning is a type of unsupervised state-action-transition-reward training mainly used to learn games right now.
Reproducibility and the Philosophy of Data with Clare Gollnick - TWiML Talk #121
Studies are not reproducible and it's not a good look for the industry. Solution: do real research, don't just throw stuff at a wall and see what sticks.
Optimal decision-making with POMDPs
POMDPs are Markov decision processes that have to deal with a partially obervable game.
Report on Geoff Hinton and his capsule networks
Bayesian Machine Learning
Intro to Bayesian Machine Learning
The cold start problem: how to build your machine learning portfolio
Build a project with an interesting dataset that took obvious effort to collect, and make it as visually impactful as possible.
Kalman filters are a simple but powerful technique for determining the most likely current state of an object in motion.
False Discovery Rates
How to reduce the noise of making multiple comparisons.
How to unit test machine learning code
Actual code examples for testing neural networks and ML algorithms yay