In the past decade or so, businesses have realized that they can extract value out of the data they collect (e.g. user data and event data) to make data-informed decisions that replace the old model of deciding best argument
Beyond simulation, applying RL in real world: Challenges and Opportunities
Standard RL offers methods for solving sequential decision making problems (typically formalized by a Markov Decision Process) that involve Maximisation of reward/minimize cost over some time horizon (and where a model of the environment may or may not be
Transfer Learning in NLP
Language has historically been difficult for computers to ‘understand’. In NLP self-supervised learning, allows us to train models with a huge amount of unlabelled training data i.e., millions of sentences from the Internet (e.g. ), which saves a lot
Exploratory Data Analysis
The greatest value of a picture is when it forces us to notice what we never expected to see. -John Tukey In the bigger picture Data-driven science, we start by collecting a data set of reasonable size, and then looking
Counterfactual policy evaluation
Making Better Decisions with Offline Deep Reinforcement Learning Introduction: In the age of big data, we have massive amounts of data available at our disposal and it would nice if we could use some of that to make better decisions
Understanding Explore vs Exploit Dilemma in Online decision-making
Online decision-making in the context of reinforcement learning involves a fundamental choice: Exploitation: Make the best decision given current information Exploration: Gather more information The best long-term strategy may involve short-term sacrifices Gather enough information to make the best overall
Understanding AlphaGo Fundamentals
AlphaGo is the first computer program to defeat a professional human Go player, the first to defeat a Go world champion, and is arguably the strongest Go player in history. This is one the major feats of AI research in
Getting Started with LISP
Developed by John McCarthy, MIT, in the late 1950s the original idea was to implement a language based on mathematical functions that McCarthy was using to model problem solving. A grad student implemented the first interpreter, and it was quickly