COM-406: Foundations of Data Science

Lectures in this course (89)

Explores the balance between exploration and exploitation in multi-arm bandit algorithms.

Covers the exploration vs. exploitation dilemma in multi-arm bandits using the Upper Confidence Bound algorithm.

Introduces contextual bandits for content selection based on different contexts.

Explores algorithms for balancing exploration and exploitation in a probabilistic environment.

Covers multi-arm bandits and distribution estimation, emphasizing the importance of robust estimators.

Explores multi-arm bandits in adversarial settings and strategies for optimizing rewards.

Explores distribution estimation methods, fitness functions, and the importance of selecting the right estimator for accurate results.

Covers methods to estimate probability distributions from samples.

Covers property testing for uniform distributions and the importance of reliable predictions.

Covers the estimation of distributions using various methods such as minimum loss and expectation.

Covers property testing against uniform distributions using the plug-in approach.

Covers the concept of distribution estimation and the optimization of parameters using different estimators.

Covers test statistics, threshold values, and lower bounds for estimating properties.

Explores distribution estimation, constraints in estimators, and competitive analysis for robust estimation.

Explores exponential families, Bernoulli distributions, parameter estimation, and maximum entropy distributions in statistical modeling.

Covers the estimation of distributions using samples and probability models.

Covers the concept of property testing using statistical methods.

Covers exponential families and maximum entropy distributions under moment constraints.

Covers the exponential family of distributions and the Ising model, explaining magnetism through dipoles and spins.

Covers binary classification, hypothesis testing, likelihood ratio tests, and decision rules.

Page 3 of 5