Chevron Left
Volver a Mathematics for Machine Learning: PCA

Opiniones y comentarios de aprendices correspondientes a Mathematics for Machine Learning: PCA por parte de Imperial College London

2,836 calificaciones

Acerca del Curso

This intermediate-level course introduces the mathematical foundations to derive Principal Component Analysis (PCA), a fundamental dimensionality reduction technique. We'll cover some basic statistics of data sets, such as mean values and variances, we'll compute distances and angles between vectors using inner products and derive orthogonal projections of data onto lower-dimensional subspaces. Using all these tools, we'll then derive PCA as a method that minimizes the average squared reconstruction error between data points and their reconstruction. At the end of this course, you'll be familiar with important mathematical concepts and you can implement PCA all by yourself. If you’re struggling, you'll find a set of jupyter notebooks that will allow you to explore properties of the techniques and walk you through what you need to do to get on track. If you are already an expert, this course may refresh some of your knowledge. The lectures, examples and exercises require: 1. Some ability of abstract thinking 2. Good background in linear algebra (e.g., matrix and vector algebra, linear independence, basis) 3. Basic background in multivariate calculus (e.g., partial derivatives, basic optimization) 4. Basic knowledge in python programming and numpy Disclaimer: This course is substantially more abstract and requires more programming than the other two courses of the specialization. However, this type of abstract thinking, algebraic manipulation and programming is necessary if you want to understand and develop machine learning algorithms....

Principales reseñas


6 de jul. de 2021

Now i feel confident about pursuing machine learning courses in the future as I have learned most of the mathematics which will be helpful in building the base for machine learning, data science.


16 de jul. de 2018

This is one hell of an inspiring course that demystified the difficult concepts and math behind PCA. Excellent instructors in imparting the these knowledge with easy-to-understand illustrations.

Filtrar por:

576 - 600 de 706 revisiones para Mathematics for Machine Learning: PCA

por Sagun S

14 de mar. de 2019

Tough one if you are new to programming or doesn't have excellent understanding of Maths

por Keng C C

30 de may. de 2020

explanations are not clear, need to refer to lots of youtube to catch up with course.

por Matan A

20 de oct. de 2019

The is a lot of gap from what the lecturer learn and what the assignments requires.

por Yuxuan W

5 de oct. de 2018

Always spending much more time on coding than needed. Same result but no credit :(

por PS

2 de mar. de 2021

Too much material covered too quickly. Needs to be split into seperate modules.

por Sethu N O G

16 de ago. de 2020

faculty must improve his teaching techniques.

I found the course less interesting

por Rafael C

7 de dic. de 2019

definitely one of the most catastrophic courses I've ever taken on Coursera...

por Sherryl S S

7 de mar. de 2021

Not enough explanation, minimum instructions, hard projects, lots of errors.

por Meraldo A

8 de may. de 2018

The course content was good; however, it was not well explained at times.

por connie

21 de mar. de 2020

I think content of first 2 weeks are disconnect with 3rd and 4th weeks

por Alexander

6 de nov. de 2019

Math for the sake of math. Too big jumps in calculations, too complex.

por k v k

30 de nov. de 2018

its a good course to learn mathematics essential for machine learning

por Rafael C

24 de sep. de 2019

The Classes didn't give the knowledge to solve the assignments.

por Shuyu Z

18 de oct. de 2019

The videos and instructions for the assignment are not clear.

por gaurav k

3 de jul. de 2019

More examples and visualization should be there to explain.

por Malcolm M

5 de mar. de 2019

Far more challenging than the first two courses.

por A. S M S H

2 de jun. de 2020

Theories should be explained more detailed.


26 de feb. de 2020

Last assignment was hell on Earth...

por Nicolas G

12 de abr. de 2021

Very bad course, poorly explained

por kirellos h

8 de abr. de 2020

This course needs more examples.

por Sean W

25 de nov. de 2019

Notebook extremely buggy

por Felipe R D

20 de ago. de 2021

No clear explanations

por Felipe M

26 de jul. de 2020

It is a shame that this course isn't taught in a favorable way, as the content it has is very interesting and valuable. I found that the instructor lacked the enthusiasm that David Dye and Sam Cooper had in the previous courses, which obviously doesn't change the content of the course but definitely makes the learning experience worse. The lectures were also quite fast-paced and not very clear, I feel that this course should have been longer as when it was time to do the graded assignments, I had very little intuitive understanding of the concepts learned. The programming assignments were also the worse of the three courses; this is a combination of what I believe to be an issue with Coursera's online programming environment and the assignments themselves. The assignments were poorly explained and usually involved skills that were not even presented in lectures, which meant that unfortunately I had to rely heavily on books from the internet and assistance from fellow peers in the forums. Apart from requiring skills that were not taught, the Jupyter Notebook was unorganized in the sense that I felt unclear about where I should edit, where I should not. The programming assignments with the previous courses in this specialization were done in a much better way, guiding us to the solution while still demanding creativity and insight into the concepts, while the ones in PCA were messy. This is really sad as this is the most programming-heavy course. Overall I am quite disappointed with this course, it is a frustrating way to end this specialization with the two amazing previous courses.

por Pedro L

25 de abr. de 2020

Having taken the other two courses for this specialization, a certain standard was defined and expected. The other two courses had solid basis explained by the professors, and the assignments reflected well from the lessons showing a lineal progression to adequate difficulty.

In this course unfortunately it is not the case, the maths and basis are explained well enough, with extra lectures and side investigations needed from the user side in order to fully understand each lecture, and then the assignments. Don't expect immediate response form mentors nor teaching staff, and neither a well thought difficulty progression. The assignments done by hand and examples taught during lectures DO NOT reflect the difficulty level on programming assignments because it is expected you already have previous experience with python (which is rather frustrating as I took this course expecting to be entry level only on this language).

TL;DR: Take the first two courses if you wanna strengthen your basis, but the last course is not recommended

por Jim F

12 de abr. de 2021

This course on PCA did not live up to my expectation from the previous two courses in the specialization. The first two courses were clearly designed to fit together, but this third course felt like it hadn't been designed to fit into this specialization. It covered material that had already been covered, and assumed other knowledge that hadn't been covered.

The teaching style was also very different: the teacher spent almost no time developing intuition with graphs or motivation with real-world problems, and instead nearly all the time was devoted to algebraic derivations. (Weirdly, the end of the final week did cover some of this; it should instead have been at the start!)

A full 25% of the course was on the abstract notion of "inner products", which was not even necessary to understand PCA. We just used the dot product.

IMO, PCA should not be its own course; it could be condensed to one or two weeks in the other courses. A much more useful third course for ML would be a Probability/Statistics primer.