Machine Learning is the basis for the most exciting careers in data analysis today. You’ll learn the models and methods and apply them to real world situations ranging from identifying trending news topics, to building recommendation engines, ranking sports teams and plotting the path of movie zombies.

Major perspectives covered include:

probabilistic versus non-probabilistic modeling

supervised versus unsupervised learning

Topics include: classification and regression, clustering methods, sequential models, matrix factorization, topic modeling and model selection.

Methods include: linear and logistic regression, support vector machines, tree classifiers, boosting, maximum likelihood and MAP inference, EM algorithm, hidden Markov models, Kalman filters, k-means, Gaussian mixture models, among others.

In the first half of the course we will cover supervised learning techniques for regression and classification. In this framework, we possess an output or response that we wish to predict based on a set of inputs. We will discuss several fundamental methods for performing this task and algorithms for their optimization. Our approach will be more practically motivated, meaning we will fully develop a mathematical understanding of the respective algorithms, but we will only briefly touch on abstract learning theory.

In the second half of the course we shift to unsupervised learning techniques. In these problems the end goal less clear-cut than predicting an output based on a corresponding input. We will cover three fundamental problems of unsupervised learning: data clustering, matrix factorization, and sequential models for order-dependent data. Some applications of these models include object recommendation and topic modeling.

This course is part of a MicroMasters program. If you complete all courses in the MicroMasters program in 2018, GE will guarantee you an interview in Boston for an internship or full-time role. Open to Massachusetts residents only.

Syllabus

Week 1: maximum likelihood estimation, linear regression, least squares Week 2: ridge regression, bias-variance, Bayes rule, maximum a posteriori inference Week 3: Bayesian linear regression, sparsity, subset selection for linear regression Week 4: nearest neighbor classification, Bayes classifiers, linear classifiers, perceptron Week 5: logistic regression, Laplace approximation, kernel methods, Gaussian processes Week 6: maximum margin, support vector machines, trees, random forests, boosting Week 7: clustering, k-means, EM algorithm, missing data Week 8: mixtures of Gaussians, matrix factorization Week 9: non-negative matrix factorization, latent factor models, PCA and variations Week 10: Markov models, hidden Markov models Week 11: continuous state-space models, association analysis Week 12: model selection, next steps

- No clarity if a particular variable is a scalar, vector or matrix

- Professor simply reads from the slides and does not add much to what forms part of the slide deck.

- No formative assessments (the only quizzes that we take are high pressure do or die quizzes)

- Many trick questions in the quizzes (for example: " Which of the following are NOT active learning strategies?")

- Ambiguous language ("This question tests a fundamental…

Following are the problems with the course:

- No diagrams

- No examples

- No clarity if a particular variable is a scalar, vector or matrix

- Professor simply reads from the slides and does not add much to what forms part of the slide deck.

- No formative assessments (the only quizzes that we take are high pressure do or die quizzes)

- Many trick questions in the quizzes (for example: " Which of the following are NOT active learning strategies?")

- Ambiguous language ("This question tests a fundamental property of the Gaussian distribution that could be considered a probability prerequisite." - I don't know if he is referring to the fact that you need to know this in order to know probabilities or does probability theory rest on this property?)

- Questions outside the textbook and lectures (i.e. that are not part of the lectures.)

- He does not move from concrete examples to an abstraction of the concept. Rather he moves in the reverse direction - moving from an abstract concept and usually stays there ... but rarely when he comes down to make it concrete, it's too late.

- In short, this course offers commoditised and generic information in an uninspired and dry teaching fashion. You can find far better lectures for most of the topics in the course on YouTube. The professor builds walls that you have to climb in order to understand what he says. He does not build bridges that help you cross the river of understanding.

If you have done other ML courses (like Andrew NG's or Nando de Freitas' courses) you will be sorely disappointed.

by
Kushcompleted this course, spending 7 hours a week on it and found the course difficulty to be hard.

There are not many courses online that provide such in-depth learning experience in Machine Learning. This course goes into some details and mathematics of the algorithms being used. It demands a good amount of time every week to understand and apply all that is being taught but that is what makes it good. It is not like many other courses that you can take and pass with minimal effort but at the end of it, it is worth spending time taking this course.

A great introduction to machine learning for those who are well grounded in the mathematics of undergraduate level. It explains the algorithm and the mathematical background of various machine learning techniques very clearly.

Only one point I did not like about this course was that the assignments were not well organized. I hope it will be improved in the future.

by
Harishcompleted this course, spending 6 hours a week on it and found the course difficulty to be hard.

Beautiful course which covers advanced concepts of machine learning. Professor Paisley covers a whole range of topics and breaks down hard concepts clearly. This course is very theoretical. There are four programming assignments which give an opportunity implement some of the algorithms learned. The only downside is that weekly assignments lack mathematical/programming rigor which can be improved in future sessions. Overall must do course for anyone interested in this topic.

Senseicompleted this course and found the course difficulty to be hard.

This course requires a solid foundation on probabilities, calculus, linear algebra and programming. Provided these prerequisites are available (anyone who is serious about the field should possess these skills anyway), the course will become an incredibly useful resource to break into Machine Learning. The only downside I found is that neural networks is not covered. A great deal of the current breakthroughs in ML are happening in this area!