Hacker News Reader

Comment by lathamcity | Hacker News Reader

I'm in the middle of the machine learning coursera course, and registered for this one as well due to interest in the material.

My one complaint is that the programming assignments weren't interesting at all. The results were interesting, but the setups were mostly given to us, and we just had to code an algorithm that was in our notes. For someone who understands the basics of linear algebra and programming, it was just a syntax challenge, and that got irritating after a bit so I stopped doing them.

I won't get the certificate for completing the course, but I have a few extra hours of free time each week to add this second course, so I'm happy. I doubt that the actual homework that Stanford students taking this course get is so easy and repetitive, though, and I'm positive they wouldn't complain about not getting to retake quizzes after getting poor grades.

Not to knock the course. I've learned a lot and the professor (Andrew Ng) does a good job.

View on HN

I've taken both, and the code is in fact not that much simpler than it was in the original class. There are, however, two huge differences: the algorithm is spoon-fed to you, and there is no math.

Firstly, think about how much more difficult the assignments would be if, for example, the steps weren't broken out and we didn't get any advice on how to vectorize. Of course, it would still be short work for anyone who (a) knows Matlab/Octave and/or (b) understands the material well, but it would also be an order of magnitude harder.

Secondly - and this is by far the larger point - the original CS 229 was really about math; the programming assignments were more of an afterthought. The lectures and homework mainly focused on the theoretical derivations and corollaries of the math that led to the algorithms. Once you'd done your bit on the math and cried to your classmates and the TA about it, you could go and implement the beautiful and extremely succinct result in Matlab.

As for my perspective on the difference, I believe it is a deliberate choice made with full knowledge of the difficulty drop. For starters, there are (with regards to homework help) no TAs in this course, so the absolute difficulty would have to decline to create an equivalent experience. More significantly, the enrollment has increased by a factor of about 700. If Stanford students had trouble with the original, you can bet that the median student in the course doesn't find it as easy as either of us does. If the goal is to generate the greatest benefit for the most people, and delivering the algorithms with a good intuition on their proper use will do so, then this course has succeeded marvelously. Of course, the smartest and most dedicated students will want more, which remains available through textbooks as well as the original course handouts (http://cs229.stanford.edu/materials.html). However, I would argue that the goal of most MOOCs (massive open online courses) should be to kindle interest and foster basic understanding, both of which the Coursera version achieves.

(slightly old) lecture videos for CS 229: http://www.youtube.com/course?list=ECA89DCFA6ADACE599

Hi,

I am also taking the course by Andrew Ng and understand your complaint that the programming assignments aren't as interesting ( from your perspective). Being quite comfortable with linear algebra, I was able to complete the assignments easily.

But when I go through the course forums, I find that for many people taking the course, the intuition behind the use of linear algebra in ML doesn't come as easy as it does for us. I think when Andrew Ng designed this online course, he must have had those people in mind also. I think he mentions it at the start of the course that it's more about understanding the concepts and the implementation details should come later. The programming exercises are designed keeping that in mind, I think.

I tried to make the programming exercises interesting for myself, by first thoroughly understanding the code that they had provided and tweaking it here and there. Once you have done that, you could apply what you've learnt on real world datasets from sources like Kaggle and see how you fare :)

> My one complaint is that the programming assignments weren't interesting at all. The results were interesting, but the setups were mostly given to us, and we just had to code an algorithm that was in our notes. For someone who understands the basics of linear algebra and programming, it was just a syntax challenge, and that got irritating after a bit so I stopped doing them.

I agree with this. The programming assignments I've done so far in the Machine Learning class are usually 5-7 matlab functions, many of which are about 2 lines of code (the longer ones might be ~10 lines of code). If you've ever done matlab/octave programming the assignments will take about 20-30 minutes and be completely unenlightening as you're literally just translating mathematical notation into matlab (which is, by design, already a lot like mathematical notation anyway). They provide entirely way too much skeleton code to learn anything from if you're not actively trying to learn. If I weren't already mostly familiar with most of the material presented in the class, I imagine I would never retain knowledge of how the machine learning "pipeline" worked or have any high-level understanding of the algorithms, because the assignments just require you to implement the mathematical pieces of each step, without ever asking you to, for example, actually call any optimization routines, or put the pipeline together.

The problem, I think, is that it would just be too difficult to do automatic grading in a way that is reasonably possible to pass if they don't turn most of the work into skeleton code. Since the automatic grading needs nearly exactly matching results, one minor implementation difference in a perfectly good implementation of the algorithm itself (e.g., picking a single parameter incorrectly, picking the optimization termination conditions incorrectly, choosing a different train/dev split, etc.) would make the entire solution completely wrong.

I'm doing the Computational Finance class via Coursera at the moment, and I've done a number of other courses previously.

I agree the programming assignments in the Finance class tend to be too simple. Most of the code is literally handed to you, you just have to understand it well enough to change it. I also understand that even that can be a major challenge if you don't have the background for it.

But I'm choosing to see the class itself as a starting point. It's a framework for my own explorations into the topics. I can do the minimum and get the minimum out of it. Or I can use what's provided as a base and go further.

The Coursera Algorithms class, for example. Writing code that got the answer was relatively easy, so once that step was done it became about optimizing the code for my own learning benefit.

It's like any educational process, you get our what you put in.

Right, you can get more out of the assignments if you try, but to me the purpose of assignments (versus passive learning - lectures, reading, etc.) is to force your brain to synthesize rather than just comprehend. The ideal assignment, then, is one that forces you to synthesize as many concepts it intends to teach as possible.

Just like you could go back and implement for yourself the skeleton code they handed you, you could also go out and implement everything in the lectures without any assignments at all. It's just that, like you said, the assignments provide a useful starting point. And I'm only saying they could be even more useful by requiring you to implement more of the complete pipeline.

The fact that an incredibly self-motivated person could learn everything there is to know about machine learning with the course as a starting point doesn't mean that it's bad to make the course more useful for a somewhat lazier or less interested person.

I've noticed this to be the case too with other courses. So for this one I've decided to implement everything in Scala (I'm currently taking the functional programming course as well. This will work well since this machine learning course requires no code submission and just questions about the results

I thought about doing it in scala too but I thought there might be issues with grading, do you know if there's an auto-grader for this course?

I took CS229 here at Stanford and I was also one of the TAs for the online version last year (I was one of 2.5 people involved with making the programming assignments).

First, the Stanford CS229 version is definitely much more difficult than what you guys had online. The focus in the actual class was on the math, derivations and proofs. The homeworks sometimes got quite tricky and took a group of us PhD students usually about 2 days to complete. There was some programming in the class but it was not auto-graded so usually we produced plots, printed them out, attached the code and had it all graded by TAs for correctness. The code we wrote was largely written without starter code and I do believe you learn more this way.

An online version of the class comes with several challenges. First, you have to largely resort to quizzes to test students (instead of marking proofs, derivations, math). There is also no trivial way to autograde resulting plots, so everything has to be more controlled, standardized and therefore include more skeleton code. But even having said all that, Andrew was tightly involved with the entire course design and he had a specific level of difficulty in mind. He wanted us to babysit the students a little and he explicitly approved every assignment before we pushed it out. In short, the intent was to reach as many people as possible (after all, scaling up education is the goal here) while giving a good flavor of applied Machine Learning.

I guess what I mean is that you have more experience than the target audience that the class was intended for and I hope they can put up more advanced classes once some basics are covered (Daphne Koller's PGM class is a step in this direction). But there are still challenges with the online classes model. Do you have ideas on how one can go beyond quizzes, or how one can scale down on the skeleton code while retaining (or indeed, increasing) the scale at which the course is taught?

I think peer-graded assignments might do the job. I am taking Gamification course on Coursera right now, and I liked peer-graded assignments a lot.

If there would be peer-graded assignments in machine learning course, I would definitely have tried them out.

Try the Learning From Data course : http://work.caltech.edu/telecourse.html A Fall run has just started (on the 2nd of oct.).

It's the same version as the course given at CalTech and is more in-depth than Andrew Ng's. There is no skeleton code for the programming assignments, answers are made through quizzes. I took the summer session and learned a lot from it.

Great. Do you get a completion certificate at the end of course?

I took the course in the spring, and found it interestng, and the programmin assignments fairly easy. This summer I took the ML course that Caltech offered, which was significantly more challenging (the homework assignments were multiple choice, but they often required writing substantial code, without any starter code.) The Caltech course is now available on iTunes U...

> The results were interesting, but the setups were mostly given to us, and we just had to code an algorithm that was in our notes.

Right; I agree. I'm not sure how they would go about making it more challenging though. They can't expect us to go out and collect data ourselves, after all. I suppose they could give us the data, then expect us to code the setup and algorithms up ourselves, but that, too, would become repetitive after a few assignments.

> Not to knock the course. I've learned a lot and the professor (Andrew Ng) does a good job.

Agreed once again. I knew nothing about machine learning before starting; now I know about neural networks, SVMs, and PCM. It's really cool how much I've learned already, for free, too!

I've also signed up for this course, but the quizzes really aren't up to par. As an example: the first quiz question was about training a neural network with too much data, and about whether or not said network would be able to generalize to new test cases. Overfitting neural networks wasn't even mentioned in the lectures; I had to rely on material from Andrew's class to answer the question correctly. This chasm between the lectures and the quizzes is likely because Geoffrey is the one creating the video lectures, but he's not the one creating the quiz questions; he is having TAs do it [1].

Nevertheless, it looks like they're responding to feedback, so hopefully it'll get better with time.

1. https://class.coursera.org/neuralnets-2012-001/wiki/view?pag...

(PCM) -> Do you mean PCA (Pricipal Component Analysis)?

I'm positive they wouldn't complain about not getting to retake quizzes after getting poor grades.

My experience is that students everywhere complain about grading. I've never been to Stanford, but I've attended and worked at several other top tier universities.

Reply on news.ycombinator.com