Bayesian methods are used in lots of fields: from game development to drug discovery. They give superpowers to many machine learning algorithms: handling missing data, extracting much more information from small datasets. Bayesian methods also allow us to estimate uncertainty in predictions, which is a really desirable feature for fields like medicine.
When Bayesian methods are applied to deep learning, it turns out that they allow you to compress your models 100 folds, and automatically tune hyperparametrs, saving your time and money.
In six weeks we will discuss the basics of Bayesian methods: from how to define a probabilistic model to how to make predictions from it. We will see how one can fully automate this workflow and how to speed it up using some advanced techniques.
We will also see applications of Bayesian methods to deep learning and how to generate new images with it. We will see how new drugs that cure severe diseases be found with Bayesian methods.
Este curso forma parte de Programa Especializado - Advanced Machine Learning
Bayesian Methods for Machine Learning
ofrecido por
Programa Especializado - Advanced Machine LearningNational Research University Higher School of Economics
Acerca de este Curso
4.6
234 calificaciones
Habilidades que obtendrás
Bayesian OptimizationGaussian ProcessMarkov Chain Monte Carlo (MCMC)Variational Bayesian Methods
Programa - Qué aprenderás en este curso
Semana
1Introduction to Bayesian methods & Conjugate priors
Welcome to first week of our course! Today we will discuss what bayesian methods are and what are probabilistic models. We will see how they can be used to model real-life situations and how to make conclusions from them. We will also learn about conjugate priors — a class of models where all math becomes really simple.
9 videos (Total 55 min), 1 reading, 2 quizzes
9 videos
Bayesian approach to statistics5m
How to define a model3m
Example: thief & alarm11m
Linear regression10m
Analytical inference3m
Conjugate distributions2m
Example: Normal, precision5m
Example: Bernoulli4m
1 lectura
MLE estimation of Gaussian mean10m
2 ejercicios de práctica
Introduction to Bayesian methods20m
Conjugate priors12m
Semana
2Expectation-Maximization algorithm
This week we will about the central topic in probabilistic modeling: the Latent Variable Models and how to train them, namely the Expectation Maximization algorithm. We will see models for clustering and dimensionality reduction where Expectation Maximization algorithm can be applied as is. In the following weeks, we will spend weeks 3, 4, and 5 discussing numerous extensions to this algorithm to make it work for more complicated models and scale to large datasets.
17 videos (Total 168 min), 3 quizzes
17 videos
Probabilistic clustering6m
Gaussian Mixture Model10m
Training GMM10m
Example of GMM training10m
Jensen's inequality & Kullback Leibler divergence9m
Expectation-Maximization algorithm10m
E-step details12m
M-step details6m
Example: EM for discrete mixture, E-step10m
Example: EM for discrete mixture, M-step12m
Summary of Expectation Maximization6m
General EM for GMM12m
K-means from probabilistic perspective9m
K-means, M-step7m
Probabilistic PCA13m
EM for Probabilistic PCA7m
2 ejercicios de práctica
EM algorithm8m
Latent Variable Models and EM algorithm10m
Semana
3Variational Inference & Latent Dirichlet Allocation
This week we will move on to approximate inference methods. We will see why we care about approximating distributions and see variational inference — one of the most powerful methods for this task. We will also see mean-field approximation in details. And apply it to text-mining algorithm called Latent Dirichlet Allocation
11 videos (Total 98 min), 2 quizzes
11 videos
Mean field approximation13m
Example: Ising model15m
Variational EM & Review5m
Topic modeling5m
Dirichlet distribution6m
Latent Dirichlet Allocation5m
LDA: E-step, theta11m
LDA: E-step, z8m
LDA: M-step & prediction13m
Extensions of LDA5m
2 ejercicios de práctica
Variational inference15m
Latent Dirichlet Allocation15m
Semana
4Markov chain Monte Carlo
This week we will learn how to approximate training and inference with sampling and how to sample from complicated distributions. This will allow us to build simple method to deal with LDA and with Bayesian Neural Networks — Neural Networks which weights are random variables themselves and instead of training (finding the best value for the weights) we will sample from the posterior distributions on weights.
11 videos (Total 122 min), 2 quizzes
11 videos
Sampling from 1-d distributions13m
Markov Chains13m
Gibbs sampling12m
Example of Gibbs sampling7m
Metropolis-Hastings8m
Metropolis-Hastings: choosing the critic8m
Example of Metropolis-Hastings9m
Markov Chain Monte Carlo summary8m
MCMC for LDA15m
Bayesian Neural Networks11m
1 ejercicio de práctica
Markov Chain Monte Carlo20m
83%
consiguió un beneficio tangible en su carrera profesional gracias a este curso
Principales revisiones
por JG•Nov 18th 2017
This course is little difficult. But I could find very helpful.\n\nAlso, I didn't find better course on Bayesian anywhere on the net. So I will recommend this if anyone wants to die into bayesian.
por AE•May 9th 2018
Challenging, but well designed course covering cutting edge ML methods. The course assumes high proficency with Tensorflow, Keras, and Python.
Instructores
Daniil Polykovskiy
ResearcherHSE Faculty of Computer Science
Alexander Novikov
ResearcherHSE Faculty of Computer Science
Acerca de National Research University Higher School of Economics
National Research University - Higher School of Economics (HSE) is one of the top research universities in Russia. Established in 1992 to promote new research and teaching in economics and related disciplines, it now offers programs at all levels of university education across an extraordinary range of fields of study including business, sociology, cultural studies, philosophy, political science, international relations, law, Asian studies, media and communications, IT, mathematics, engineering, and more.
Learn more on www.hse.ru
Acerca del programa especializado Advanced Machine Learning
This specialization gives an introduction to deep learning, reinforcement learning, natural language understanding, computer vision and Bayesian methods. Top Kaggle machine learning practitioners and CERN scientists will share their experience of solving real-world problems and help you to fill the gaps between theory and practice. Upon completion of 7 courses you will be able to apply modern machine learning methods in enterprise and understand the caveats of real-world data and settings.
Preguntas Frecuentes
¿Cuándo podré acceder a las lecciones y tareas?
Una vez que te inscribes para obtener un Certificado, tendrás acceso a todos los videos, cuestionarios y tareas de programación (si corresponde). Las tareas calificadas por compañeros solo pueden enviarse y revisarse una vez que haya comenzado tu sesión. Si eliges explorar el curso sin comprarlo, es posible que no puedas acceder a determinadas tareas.
¿Qué recibiré si me suscribo a este Programa especializado?
Cuando te inscribes en un curso, obtienes acceso a todos los cursos que forman parte del Programa especializado y te darán un Certificado cuando completes el trabajo. Se añadirá tu Certificado electrónico a la página Logros. Desde allí, puedes imprimir tu Certificado o añadirlo a tu perfil de LinkedIn. Si solo quieres leer y visualizar el contenido del curso, puedes auditar el curso sin costo.
¿Cuál es la política de reembolsos?
¿Hay ayuda económica disponible?
What background knowledge is necessary?
Course requires strong background in calculus, linear algebra, probability theory and machine learning.
¿Tienes más preguntas? Visita el Centro de Ayuda al Alumno.
Coursera brinda acceso universal a la mejor educación del mundo, al asociarse con las mejores universidades y organizaciones, para ofrecer cursos en línea.
© 2018 Coursera Inc. Todos los derechos reservados.
