Zurück zu Machine Learning: Clustering & Retrieval

4.6

1,766 Bewertungen

•

301 Bewertungen

Case Studies: Finding Similar Documents
A reader is interested in a specific news article and you want to find similar articles to recommend. What is the right notion of similarity? Moreover, what if there are millions of other documents? Each time you want to a retrieve a new document, do you need to search through all other documents? How do you group similar documents together? How do you discover new, emerging topics that the documents cover?
In this third case study, finding similar documents, you will examine similarity-based algorithms for retrieval. In this course, you will also examine structured representations for describing the documents in the corpus, including clustering and mixed membership models, such as latent Dirichlet allocation (LDA). You will implement expectation maximization (EM) to learn the document clusterings, and see how to scale the methods using MapReduce.
Learning Outcomes: By the end of this course, you will be able to:
-Create a document retrieval system using k-nearest neighbors.
-Identify various similarity metrics for text data.
-Reduce computations in k-nearest neighbor search by using KD-trees.
-Produce approximate nearest neighbors using locality sensitive hashing.
-Compare and contrast supervised and unsupervised learning tasks.
-Cluster documents by topic using k-means.
-Describe how to parallelize k-means using MapReduce.
-Examine probabilistic clustering approaches using mixtures models.
-Fit a mixture of Gaussian model using expectation maximization (EM).
-Perform mixed membership modeling using latent Dirichlet allocation (LDA).
-Describe the steps of a Gibbs sampler and how to use its output to draw inferences.
-Compare and contrast initialization techniques for non-convex optimization objectives.
-Implement these techniques in Python....

Jan 17, 2017

Excellent course, well thought out lectures and problem sets. The programming assignments offer an appropriate amount of guidance that allows the students to work through the material on their own.

Aug 25, 2016

excellent material! It would be nice, however, to mention some reading material, books or articles, for those interested in the details and the theories behind the concepts presented in the course.

Filtern nach:

von Rajkumar K

•May 27, 2017

Clustering & Retrieval was a lot tougher compared to courses on regression & classification because the match concepts behind this course were too complex. Nevertheless Emily tried to make this course as intuitive as possible

von Gilles D

•Aug 12, 2016

Still a very good course.

Week 4 was very tough. The general concept can be understood from a 10,000 feet altitude but the lesson and programming assignment need to be reviewed, maybe with a slower step by step example.

As some other student mentioned, it was... "brutal".

Other than that looking forward to the next course in the specialization!

von Usman I

•Dec 29, 2016

I am taking all courses in the specialization, and this is my fourth. I have been having a great time with materials by both instructors so far, until I came to week 5 of this course.

Despite repeated viewing, my understanding of LDA is non-existent. The first section is fine, but starting from "Bayesian inference via Gibbs sampling," for me at least, the method of instruction has gone off a cliff.

I strongly suggest soliciting feedback from learners that narrowly targets the material of this week 5 to determine if it's just me or if this is a wider problem. If it is the latter, perhaps it is time to redesign the lessons of this week.

von Siva J

•Feb 26, 2017

Good and deep dive into ML!

Absolutely disappointed that the course was delayed and the promise to take it through Course 5 and Capstone Project didn't come through.

Not at all happy with that!!

von Koen O

•Aug 27, 2017

I liked it a lot

von Mehul P

•Sep 11, 2017

Nice explanation on clustering methods.

von Michele P

•Sep 02, 2017

Advanced course. The material taught in this course is more advanced compared to Regression and Classification courses. You have to invest more time in respect to the previous courses. For some topics (LDA and hierarchical clustering) I had to look for other sources in order to understand the concepts properly. However, this course is a good introduction to clustering and retrieval.

von Galen S

•May 08, 2017

I liked the slides.

von Maxence L

•Dec 15, 2016

Comme les précédents dans de cette spécialisation, ce cours est très riche et donne les clés pour utiliser des outils complexes et puissants. Toutefois, un peu plus de détails sur certains aspects, notamment théoriques, pourraient améliorer la compréhension de certains chapitres plus techniques.

von Marco A d S M

•Oct 20, 2017

As explicações poderiam ser um pouco mais detalhadas neste curto. Tive certa dificuldade em alguns conceitos apresentados, mais do que nos outros cursos.

von Jack B

•Mar 04, 2017

Should use pandas instead of Graph Lab Create

von Christopher M

•Jul 01, 2019

Doesn't go quite as deep into the details as some of the other Machine Learning courses from the University of Washington do. Overall though, the course covers a LOT of ground. and provides exposure to many different topics.

I would have liked to have seen an Optional section on the derivation of some of the math that we were given functions for on the Expectation Maximization section. The models in the hierarchical clustering section take longer to fit than is necessary for a course like this (more than 40 times as long as the instructions say it should take), maybe a larger tolerance for convergence should be specified?

von Baubak G

•Jul 11, 2018

Need more details in the coarse. I think many of the topics need more working on, and are not sufficiently described.

von Oliverio J S J

•Jun 20, 2018

Some of the contents of this course are interesting, but it seems that this course has been very affected by the changes that forced the cancellation of the last two courses of the specialization. Apparently, they had to redo it and there are even two fake weeks (the first one and the last one). It is a pity that they did not want to spend more time to reorganize it.

von Alexander S

•Aug 07, 2016

great course, but module 4 lacks a bit in structure. hard to follow. without the forum, it would not be possible to make it in time.

von Piotr Ś

•Feb 15, 2017

Dependence on GraphLab technology is a big minus. The lectures are poorly balanced in terms of difficulty. Apart from that - interesting course, I'm glad I took it.

von Aayush G

•Nov 10, 2016

This specific course traded off depth and detail for breadth of topics. Too many ideas were quickly described and not really built up to my liking.

von Adrien S

•Oct 07, 2016

Feels like this course in the specialization was a bit rushed, compared to the first 3 courses. It had 2 modules (first & last) that were more like placeholders and the middle 4 modules went from concept to the maths behind the algorithm very quickly. It needs a bit of work on expanding the course and presenting a bit more slowly. Having said all that, the concepts and algorithms taught are very interesting and a good first step into the unsupervised learning section.

von Kayvan S

•Feb 15, 2018

Great course but I think the workload could be spread across the weeks more. Also, I had a lot of trouble with the sklearn toolkit (probably due to installation issues.).

von Valentina S

•Aug 16, 2016

Interesting content but explanations are less clear with respect to the other courses of the ML Specialization

von Xiaosong L

•Aug 31, 2016

the homework is getting easy

von Saeed S T

•Sep 08, 2016

Overall a good and useful course, however:

A) They could do a much better job regarding LDA, standard Gibbs sampling, and Bayesian model and inference. Many slides on these 3 topics only contained some text and the instructor tried to "verbally" visualize the related important concepts. Hence not a good use of a video session.

B) Week 1 and the 1st half of Week 6 were redundant.

C) It would be much better to have a 7-week course with more topics and may be with some optional videos on Bayesian model, HMM.

von Volker H

•Jul 18, 2016

please rework in particular week 5, part 2

von Michael L

•Mar 18, 2017

slightly repetitive of classification course with no real use-case value except lots of math..

von Diego T B

•Aug 28, 2016

The retrieval part of this course is great, it deserve five starts. The clustering part was going well until it reached LDA.

The LDA module is very poorly covered, and also very hard to understand. I had to watch the videos more than two times to try to figure out what was LDA, and a Quora article posted in the Forum could explain it much better.

Then we get to the Hierarchical Clustering module, which was the most poorly module in all this specialization. There is only one video talking about HMM models, and Markov Chains deserve at least one week to even get started with it. And to complete, there is just one Assignment with only 3 questions.

The specialization was going perfect until now. I am very disappointed with this course. I hope the last two courses are much better covered and not just ran over like this this one was.

Coursera arbeitet mit erstklassigen Universitäten und Organisationen zusammen, um Online-Kurse anzubieten und dadurch universellen Zugriff auf die weltweit beste Ausbildung zu ermöglichen.