Course

PHYS-467: Machine learning for physicists

Lectures in this course (106)

Variational Methods in Statistical Physics

Explores variational methods in statistical physics and the Bethe Parisi Mézard approach.

Random Field Ising Model

Covers the Random Field Ising Model, discussing the concept of disorder and the process of finding the ground state.

Regression: High Dimensions

Explores linear regression in high dimensions and practical house price prediction from a dataset.

Optimizing Loss Functions: Gradient Descent Variants

Explores early stopping in gradient descent and variants for faster optimization.

Gradient Descent and Linear Regression

Covers stochastic gradient descent, linear regression, regularization, supervised learning, and the iterative nature of gradient descent.

Linear Regression: Regularization

Covers linear regression, regularization, and probabilistic models in generating labels.

Self-Organization in Complex Systems

Discusses self-organization in complex systems and emergent properties.

Linear Regression: Regularization Overview

Explores linear regression fundamentals, emphasizing the importance of regularization techniques to enhance model performance.

Belief Propagation: Key Methods and Analysis

Covers Belief Propagation, a key method for both analysis and algorithm.

Generalized Linear Regression

Explores generalized linear regression, logistic regression, and multiclass classification in machine learning.

Multiclass Classification

Covers the concept of multiclass classification and the challenges of linearly separating data with multiple classes.

Back to Linear Regression

Covers linear regression, regularization, inverse problems, X-ray tomography, image reconstruction, data inference, and detector intensity.

Bethe Free Entropy

Covers the computation of Bethe free entropy and the interpretation of messages between variables and factors.

Regression Again: Exercise 3.1

Covers exercises on regression, including linear and polynomial regression, high dimensions, and real data analysis.

Information Theory: Basics

Covers the basics of information theory, entropy, and fixed points in graph colorings and the Ising model.

Singular Value Decomposition

Explores Singular Value Decomposition and its role in unsupervised learning and dimensionality reduction, emphasizing its properties and applications.

Inference: Exercise 3.2

Covers the topic of likelihood estimation and maximum likelihood estimation in inference.

Principal Components Analysis

Covers Principal Components Analysis, a technique for dimensionality reduction and gene type analysis.

Fixed Points in Graph Theory

Focuses on fixed points in graph theory and their implications in algorithms and analysis.

Bayesian Inference: Estimation & Demystification

Covers the concepts of Demystification, Estimation, and Bayesian Inference in the context of Bayesian statistics.