I would've loved to give this course 5 star. I am very sad because i can't. The course material seems to be old and not exact when it comes to installation of softwares and packages like turicreate, jupyter notebook, anaconda etc. As a newbie i had a very hard time installing and getting a running ML environment. These should have better treatment in week 1, otherwise most student will loose hope and might not continue. But overall this was a great course. Thanks and Love for Carlos and Emily. <3

I thought it was a good course, interesting to see the perspective of some different ML methods that can be used. I think the biggest take away from the course is to use data and play around with it and the API's that are out there just to get a base understanding of how to play with them. Usually things have very good documentation that can be used as a help. I would recommend the course for people who want to just see what ML is about and play around with some python notebooks while doing it.

This is a nice entry-level course on Machine Learning. Emily and Carlos show their enthusiasm for ML and engage students in the subjects they are teaching.

The video lectures have a nice format in which you can see the professor while he/she is explaining the presented slides. This makes the course feel less online and more personal.

The course makes special emphasis on the practical side of the presented material. All the modules have a practical part in which you practice the theory presented.

Simple materials and clear explanations on concepts. A good course to begin with. However, students will be using graphlab (a python package for machine learning) in the entire specilizaiton instead of other common packages like pandas and scikit-learn. Beginners may feel a bit hard to implement some of the algorithms using those common packages. Also, if you choose to use scikit-learn instead of graphlab for the assignment, your answers may differ from which they use to grade the assignment.

This is a great introductory Machine Learning Course. The required knowledge is almost nothing, so it starts really basic, but got more interesting on the last weeks.

It focus on a practical approach. You will use GraphLab, a powerful tool that allows to easily start applying ML algorithms on real cases, even with just a basic theoretical understanding about why this works.

It made me think a lot how can I take advantage of this tools, and I'm excited about start working on some new projects.

This was a great overview into the history and current techniques in ML. I am an engineer in process control, so there are areas where I see this could be used. I like the fact that NN's found a place finally which is working well. Obviously the purpose of the course was not to delve too far into the maths, and as such "sells" the software which is a tad costly. Point however is the purpose of the course met the intent, and I found it really interesting and worth the time ! Thanks for this

The material in this course is very interesting. I feel comfortable with the concepts and algorithms. I am definitely prepared to utilize these skills in an entry-level manner - it will take some hands-on practice with real datasets to build expertise, understand the nuances of these approaches and expand my knowledge base. I recommend a decent level of comfort with programming. I completed the Python for Everybody specialization, but still struggled with the programming in this course.

I think It was a very useful course. The case study approach is a very efective technique to understand and practice the underlying concepts. However, i'd like to highlight maybe this quality might be affected because the explanations are using graphlab but the reality is the use of turicreate. Other thing: turicreate for windows is very restrictive, it would be good idea facilitate the installation and usability for this OS. Congratulations for the course and spec, I like it so far.

This course covers all basic machine learning tasks, including the new emerging technology like deep learning. The programming assignments are very interesting, face to real world applications, have an appropriate difficulty and are with good written guide. However, tools used in the course are SFrame and Dato Graphlab. If you are using popular toolkits including scikit-learn and pandas, you will be in trouble to set up the corresponding random seeds and fail to get the correct answer.

Good course to review with higher level application of Machine Learning.

The entire course needs to be updated and could easily be more concise.

Wish it was based on Python3.x, numpy, pandas or sci-kit, etc. What would be cool is a more database

oriented application so for example an AWS or some database integrated way (a more modern approach ) to implement, integrate, and deploy their case studies for linear regression, linear classification/classification, clustering, NLP etc,, etc.

A great introduction to machine learning with applications. However, there is still a small issue is that Graphlab is used instead of more commonly used Numpy and Pandas. I understand that Graphlab does have a great advantage in terms of not having to store large data into memory for some applications, but not directly learning tools that are actually used in the industry is still kind of a pain.

Also, RIP for the last 3 courses that never had their chance come into the specialization.

Muy buen curso que me introdujó un poco en este mundo del Machine Learning. La unica razón por la que no le doy un puntuación perfecta es porque creo que deberian actualizar los videos usando la libreria Turicreate que es la que plantean en los notebook, puede generar confusión ver como el docente usa graphlab y en los notebook usan una libreria totalmente diferente.

Aparte de eso fue una muy bonita experiencia y le agradezco a mi universidad por notificarme de estos cursos gratuitos

I think the course was great! the only thing I wish for that would have improved the course:

1) some more optional, more difficult questions, just to test and push our understanding

2) if it was easier to do using say sci-kit learn. I only had access to a windows at the start of this course, which made it very difficult to follow. Just providing .csv files in place of .sframe files at every instance would have made life easier at that point and made the course more accessible.

Very well done and gives a very fast and intense introduction to machine learning. I thought that I was prepared to be at the intermediate level but it was very challenging. What I am disappointed about was the quiz questions. My programming background helped with me with those related questions but the conceptual questions need review. Since there are so few questions, the 80% passing is a difficult mark to hit when two questions are not covered by the course material.

This is great course to start off with the basic of Machine Learning and data analytics in general. I would have preferred if they used the universal libraries in Python (MatPlotLib and Pandas) instead of GraphLabs as these coding skills aren't transferable outside this course.

But I really liked how we explored the data first in all courses to give us a sense of what it is and how it looks. Even the questions in the quizzes required some data retrieval skills.

This is merely a starter course that gets you acquainted with the basics of ML using Python. I hope to see the instructors cover algorithms in greater detail in the next few courses in the specialization. Also, they focus exclusively on Dato's Graphlab - that's a worry. I was hoping to be introduced to NumPy, SciPy and other python libraries for ML. They could have at least spent some time demonstrating why they thought Graphlab was a better alternative.

Instructors are enthusiastic and interesting to watch; the course content is at the appropriate level of abstraction to build curiosity without getting lost int he weeds. Assignments (except for deep learning) do require some coding knowledge, but script kiddies like myself can pass. Taking one star off because of the strong emphasis on using the GraphLab environment rather than standard FLOSS libraries in Python (pandas, scikit-learn, etc.) or R.

The videos are great! The ourse is well structured and introduces gradually the complexity. Unfortunately, the exercises requires the use of a specific library, instead of scikit-learn and numpy. Although this library is a great pedagogic tool, it doesn't work well with google colab and Jupyter notebook, especially when you need to display a graph or a picture. Because of this, I was not able to answer to few questions, and I to rely on my luck.

Solid overview of the core machine learning methods. Lectures are full of nice clear examples and explanations that provide enough detail to be useful without getting lost in the weeds. The only downside is the use of GraphLab, which is great because it is simple and powerful, but also problematic because it is proprietary for business. It would be a five start course if it also had complementary Python notebooks that used public packages.

Um ótimo curso introdutório. Ensina sobre os tópicos fundamentais das áreas do aprendizado de máquina unindo o um conhecimento básico sobre os próprios algoritmos e a aplicação em casos reais.

O curso não aplica o conhecimento sobre os detalhes de cada algoritmo pois isso será feito nos cursos posteriores. Usa-se o pacote de desenvolvimento com funções pré-definidas para cada método e o que se pratica em Python é o manuseio da informação.

its a basic course gives the overview of everything you need to learn in deep learning. it's would be more better if they teach this in more used library like tenser flow or keras would be lot more better.

ye, one more thing i faced problem with the quiz where you need to write a function to do certain task.it would be more helpful if you mention more clearly what parameters to pass in those function which we need to create .

thank you

Very good hands-on approach to practical machine learning. However, as the models get more complicated, some of the material isn't really explained very well, specially the last module about 'deep learning' is quite superfluous. I also wish, the course favored providing a bit more mathematical intuition over just graphlab usage. All things aside, if you are really curious about machine learning you should take this course. Be hungry!

Excellent overview of important Machine Learning concepts. There were some technical difficulties and reversed steps that were a problem at some points. If you are taking this course you can overcome those problems but I recommend the course creators take their own course to make sure everything is logically ordered and technically executable. :) Overall great course, though (content = 5, technical issues = 4). Thanks!

Course provides very good overview on ML methods and applications - the Python programming exercises are very hand on, in some situations I was missing a little bit of the theoretical background which I than looked up in other sources (e.g. what is the logistic classifier that was used and why is it useful). But of course this was only the foundations course and it provided good basis for the coming specializations.

Some details / variation about the "features" collection (training / test dataset) i.e. how could we build various "features" to tune our ML algorithm perform better prediction. Though there is a lecture on the impact of "features" on ML algorithm is part of the course; however, it would have been better to look at steps of building various types of "features". For me the "deep features" on images are still blurry.