Hacker News new | past | comments | ask | show | jobs | submit login
Show HN: HN.Academy – Top online courses recommended by Hacker News users (hn.academy)
773 points by yaj54 on Jan 3, 2019 | hide | past | favorite | 85 comments

Going to recommend the most widely taken course at Yale at the moment, which fortunately is now available as a MOOC on coursera, "The Science of Well Being": https://www.coursera.org/learn/the-science-of-well-being

This is a positive psychology course based on work by Seligman et al.

Would this be Seligman of "none of my positive psychology results replicate" fame?

Do Positive Psychology Exercises Work? A Replication of Seligman et al. (2005)

Results: Repeated measures analyses showed that the PPEs led to lasting increases in happiness, as did the positive placebo. The PPEs did not exceed the control condition in producing changes in depression over time.

Conclusions: Brief, positive psychology interventions may boost happiness through a common factor involving the activation of positive, self-relevant information rather than through other specific mechanisms. Finally, the effects of PPEs on depression may be more modest than previously assumed.


Thanks for the link. The four treatment conditions look REALLY similar to me (see pages 384-285). They're all exercises that make you think about good things in your life:

Expectancy control (early memories): “[...]Every night over the next week, set aside about 10 minutes before bed [...] to log on to this website to write about an early memory.”

Positive placebo (positive early memories, in addition to rationale above): “[...] Every night over the next week, set aside about 10 minutes before bed [...] to log on to this website to write about an early positive memory.”

Three good things (Seligman et al., 2005): “[...] log on to the website daily for seven days to list three things that went well on that day and why they happened.”

Using signature strengths in a new way (Seligman et al., 2005): “This exercise consists of two parts. You will take a questionnaire that gives you feedback about your strengths. This will take about 45 minutes. The next day you will be asked to use these strengths in new ways every day for one week [...]”

So regardless of whether the specific exercises are uniquely useful, it seems like it's valuable to make yourself focus on the positive stuff.

My point exactly.

Does this mean you're not signing up for the course? :D

Could you provide more information or sources about that?

This list mostly passes the gut check: the courses near the top are almost universally high quality, and quite reputable.

Now, why did I say "mostly" and "almost"?

Two reasons:

1. The best technical courses are on Udacity. You may disagree but I think enough people would agree that it ought to be somewhere on the list.

2. Learning to learn is popular but its kind of vacuous. (You may disagree, and this time I may be in a tiny minority who hold this view). So I think the sentiment reader might need tweaking.

Still, all in all, a great resource. I am bookmarking it and am glad you did this.

I mostly agree. However, the rankings surface the "most recommended" courses not the "best". "Most recommended" being more objective (easier turn into a number) than "best".

Udacity is the 3rd most recommended course provider after Coursera and edX. Those courses will make it into v2.

It seems to me that the comments on the learning to learn course generally read like positive recommendations, so I think HN disagrees with you there [0]. But I do want to experiment with different sentiment analyzers.

[0]: https://hn.academy/coursera.org/learn/learning-how-to-learn/

Although I can see why a technically-oriented person might say L2L is vacuous, I do disagree. It's not dense or technical, but I got two concrete insights and practices from it (and I don't think I even bothered to finish it):

1. The pomodoro technique really is useful if you're a hopeless procrastinator like me. It's saved my butt numerous times.

2. The explanation of the value of alternating between focused intense thinking and diffused thinking also helped clarify why unfocused thinking is genuinely productive (of course, as long as it's mixed with focused thinking). I've used this to work around mental blocks when I got too stuck in the focused mode.

I agree with you. These are useful concepts, but I've read 5 minute blog posts that convincingly explained these. Spreading this instruction over days/weeks seems to make my point.

That's a problem with a lot of content, right? Specifically practical concepts. Did you ever find yourself thinking the same things about your entry level college classes if you had been lucky enough to attend any?

You can read a blog post about building bicycle wheels in five minutes and get most everything you'd gain from going to a weeklong course, but the difference is all in how you'd engage with it. A longer discussion, open area for questions, and building familiarity with the material drive the skill home whereas that 5 minute blog post might as well disappear into the depths of memory the second you navigate elsewhere.

I took the course and while I agree that I wanted more depth, it certainly wasn't "vacuous" given the context. And it should be noted, the materials of the course follow some of the concepts taught within - they're doing their best to make sure you learn their basic concepts of how to learn.

The "I could've learned that in a 5 min blog post" applies to basically anything and is condescending, arrogant, and does not warrant a discussion.

Learning How To Learn is an entry level way to get from practically nothing (everyone who can comprehend the course has some fundamental basics; e.g. spaced repetition we learn at elementary school in The Netherlands) to a more productive learning environment. It does that by teaching you techniques via theory, practice, and ultimately a quiz about the material. After you completed the course, you will understand the techniques, have familiarized yourself with each of these, and you can likely apply them. It is also likely you found a preferred technique (or found a confirmation that your previous preference is the best one for you). If you already know these techniques (such as spaced repetition, focused & diffused mode, pomodoro technique) then the course has little value. You could still help people on the forums (explaining something has great value for yourself as well), or learn a thing or two from the forums or Friday mailing. If people don't know how to learn then all the other courses on Coursera and other MOOCs are useless. The popularity of the course and the positive reviews underline its usefulness.

You can find naysayers for every course. For example, I'm a naysayer of the Cryptography I course in a way. I recommend it (as far as I finished it; the first 2 chapters I did) because for me it gave me gigantic headaches and migraine. It isn't for me, I'm just not intelligent enough for it combined with the headaches and migraine it didn't work out and I had to call it. YMMV.

> The "I could've learned that in a 5 min blog post" applies to basically anything and is condescending, arrogant, and does not warrant a discussion.

You've completely missed the point and even failed to assume that the OP was well-meaning.

The whole point of making it clear that the content of an online course can easily be learned by reading a blog post on the subject is to warn potential users that they risk wasting their precious time watching hours of videos consisting mostly of useless filler when spending 5 minutes reading a blog post would get you the same results.

Frankly, this is a major and all too frequent problem witg MOOCs. Sadly, some courses are obvious exercises on how to stretch close to no content into hours of video.

Personally I would hate to waste hours of my personal time unnecessarily, thus I appreciate these warnings.

It makes no sense to make such a point about this specific course when it can be made about any course. Additionally, it is not possible to get on the same level via a 5 minute blog post compared to completing the course. Such counts for most if not all courses.

> Personally I would hate to waste hours of my personal time unnecessarily, thus I appreciate these warnings.

These warnings can be made with about any online content. It boils down to "I found it a waste of time"; data suggests the course was NOT a waste of time however. I appreciate such data over anecdotal evidence.

> It makes no sense to make such a point about this specific course when it can be made about any course.

The point is that the assertion was made regarding this specific course but not any other. You're the one trying to put up a strawman by forcing the assertion to cover other courses, when nothing of the sort has been made.

>Additionally, it is not possible to get on the same level via a 5 minute blog post compared to completing the course. Such counts for most if not all courses.

I'm not familiar with the course, but I'm very familiar with other MOOCs from edx, coursera, and udemy. Based on my personal experience I can say that some courses, particularly technical courses, are indeed packed with useless filler content, and the same learning experience can be had by reading a small blog post covering the subject.

This assertion is not uncommon, and unfortunately is a recurring problem with MOOCs.

Have you considered the background of a person talking about such a course? They might have read more than you about the topic, so they reasonably feel that the course brings nothing new.

The problem is that it doesn't bring anything new to them, but they forget that other people don't have the same knowledge as they do. This is called the course of knowledge, where you forget what it was like before you knew something.

What I can say about "Learning how to learn" specifically is that it covers how to encode information for better retention, what strategies to use to increase long-term recall (there are 3 and they're unintuitive), how to use chunking, how to build habits.

Sure, you could cover these topics in 5 minutes, but you wouldn't do them justice. Especially for someone who's encountering them for the first time.

> The best technical courses are on Udacity.

Is that universally agreed on? The Udacity courses I had looked at (full stack development, AI for self driving cars) were good and insightful, but somewhat half-baked, and not as good as edX or Coursera.

I'm not a fan of Udacity, mainly because I paid them $800 for a single course (the first in the self-driving car series) and then they blocked all my access to the course content because I didn't finish the projects in the allotted time. I was hoping to finish later because my work schedule changed, but they were completely inflexible. I should have read the fine print. The course content was OK, but I don't think they have great policies.

I was frustrated with Udacity for the same reason, plus, their ostensible expertise is way overstated [1]. The entire reason I'm paying for a course is to be able to bounce arbitrary questions off experts. They provide people who are still in the fake-it-till-you-make-it stage that's so popular.

[1] Earlier comment where I asked a basic question that went beyond the understanding of the literal face of the course. https://news.ycombinator.com/item?id=18596450#18601298

I lost access to some materials years ago too. It was a great program but I got hired and it no longer made sense to keep learning python. The switch to fixed terms is a bit off putting for me

Ah, it was so obvious and strong to me (3 certificates and many courses) that I didn't think this was debated.

I would be interested (if you have the energy) to see which Coursera and Edx courses you preferred to the Udacity courses you liked.

A while back I wanted to learn about SLAM, computer vision, robotics, drones, etc. I did all the following courses which had significant overlap in materials:


https://www.coursera.org/specializations/robotics (also available at edx as https://www.edx.org/micromasters/pennx-robotics)


And I would rank them in the order of edx, coursera, and udacity last.

It's not totally fair because the best comparison for udacity should be their self driving car nanodegree but it doesn't let you audit for free and I don't care for certificates. From the few udacity courses I did do, I felt their videos are too short and triggers my ADHD to go do something else after watching each minute or two long video. Edx/coursera felt a lot more like university lectures and felt more rigorous in comparison.

Agreed that Udacity's AI for Robotics was a weak course. I highly recommend some of Udacity's recent paid-only courses on robotics:

Self-driving car engineer nanodegree (not the one with "intro" in its name) https://www.udacity.com/course/self-driving-car-engineer-nan...

Robotics software engineer nanodegree https://www.udacity.com/course/robotics-software-engineer--n...

Flying car and autonomous flight engineer nanodegree https://www.udacity.com/course/flying-car-nanodegree--nd787

Just as a sample, this is one of many projects I completed as part of the self-driving car engineer nanodegree. My code controls a car driving on a highway with other cars.


I found edX to be pretty lacking if you wanted to learn something advanced that's not quite "flavor of the day:" no higher level pure math classes, very few algorithms classes past intro level, very little material on digital design other than the MIT 6.002 sequence. Tons of choices for machine learning, Python, and AP calculus though.

There’s a reasonably extensive MicroMasters course on Algorithms and Data Structures. How would you rate it? Comparable to finishing Skiena? CLRS? Concrete Mathematics?


What You'll Learn:

Understand essential algorithmic techniques and apply them to solve algorithmic problems Implement programs that work in less than one second even on massive datasets Test and debug your code even without knowing the input on which it fails Formulate real life computational problems as rigorous algorithmic problems Prove correctness of an algorithm and analyze its running time

Courses Algorithmic Design and Techniques Learn how to design algorithms, solve computational problems and implement solutions efficiently.

Data Structures Fundamentals Learn about data structures that are used in computational thinking – both basic and advanced.

Graph Algorithms Learn how to use algorithms to explore graphs, compute shortest distance, min spanning tree, and connected components.

NP-Complete Problems Learn about NP-complete problems, known as hard problems that can’t be solved efficiently, and practice solving them using algorithmic techniques.

String Processing and Pattern Matching Algorithms Learn about pattern matching and string processing algorithms and how they apply to interesting applications.

Dynamic Programming: Applications In Machine Learning and Genomics Learn how dynamic programming and Hidden Markov Models can be used to compare genetic strings and uncover evolution.

Graph Algorithms in Genome Sequencing Learn how graphs are used to assemble millions of pieces of DNA into a contiguous genome and use these genomes to construct a Tree of Life.

Algorithms and Data Structures Capstone Synthesize your knowledge of algorithms and biology to build your own software for solving a biological challenge.

LHTL is important for instilling a growth mindset in students and teaching effective science-backed learning techniques. If you know the 'why' behind an effective study method, you're more likely to continue to stick with it.

Two interesting takes...

Learning how to learn is a vital part of this list because online courses are notorious for having people burn out and lose interest so it's especially important that these people can stick with a course using actual techniques.

Why would someone burn out more with online courses (compared to e.g. university). Do you have any sources for this claim? I am very interested in knowing more about this!

I don’t know about burnout, but lose interest seems real. I graduated from MIT having dropped perhaps 2 or 3 courses. I’ve “dropped” at least 10x that number on EdX and Udacity.

I think it’s mostly a matter of engagement and commitment (no real penalty to dropping a MOOC and it didn’t cost much, if anything) and not that burnout is any worse. IOW, the rate of loss of interest might be no higher, but the initial level of interest might be lower, putting you closer to the threshold of dropping out.

Full time university is also the last time I had no other life obligations...

> I’ve “dropped” at least 10x that number on EdX and Udacity.

I think many online learning places are making a large mistake by not letting people sample the first 3 or 4 lessons.

I too have signed up for and left a rather large number of MOOC style courses simple because signing up for the course is often the only way for me to see if I enjoy the course material, the teacher, and the teaching style—all three of which vary wildly course by course on every site.

It seems to me they would be getting much more accurate numbers on completion if they didn’t hide the courses behind so much mystery. It would also have the benefit that we wouldn’t feel like we’ve “dropped” so many courses.

And also like you, I didn’t drop many courses at uni, but there was also more information available about the exact course material and there was much more consistency of teaching style. Also many were required for a degree and getting the degree was my primary purpose at the time, MOOCs however are more of a free time fun time thing for me so I’m much more picky.

I do agree that the "free lick" would lead to dramatically lower drop out rates, but think that the sites probably care more about "active users" or "course enrollments" as a primary metric of growth/market share.

It's a bit late now but i want to clarify I did mean "lose interest" not "burn out".. Used the wrong phrase completely.

I agree that learning how to learn is a vital skill, and sticking to stuff is an important habit.

However, my experience was that this class didn't effectively develop that skill or the habit for me.

> 2. Learning to learn is popular but its kind of vacuous. (You may disagree, and this time I may be in a tiny minority who hold this view). So I think the sentiment reader might need tweaking.

I agree. I thought that it would teach me something revolutionary but it didn't. Granted, some of the things they teach are useful.

I think the course is just a victim of some people hyping it up too much.

As someone who had taken the course and let some time pass, I thought the same. A stringing together of common sense ideas to learn better. Yet, to provide better context for another comment, I searched online for the course notes and came up with two excellent links. I realized there was more to the course than what I had remembered. Linking them here hoping it will help.



Hey HN,

Check out HN.Academy and let me know what you think. It's the result of mining the HN archives for references to online courses and then ranking them and displaying all references in one place.

Ranking currently takes into account HN stories (points) and comments (sentiment, karma, estimated points).

I had started with a much broader set of course providers, but Coursera ended up swamping the rankings of pretty much all the other providers. edX is also in the rankings with a few courses. I plan to add more providers but as of now none of them will impact the top rankings.

If you were looking to start a new learning endeavor in the new year... here ya go.

Best, Yaj

Interesting, wonder if that's because Coursera is perceived and thought of as better in general or is it due to other reasons such as general popularity. May make sense to have some kind of normalization to unbias it if the latter.

too bad, you weren't able to rank these by difficulty and/or skill level.

even without that, this is an awesome list. thanks.

You're missing fast.ai for machine learning. It's been mentioned or linked 700+ times on HN.

I'll need to take the fast.ai course first so that I can upgrade my course-extraction-from-text code from regexes to deep learning. ;-)

agree - fast.ai is a fantastic resource - i wish everything i want to learn about was taught in it's very practical style

I'd throw this resource in the ring since it has the most material out of anything on the list:


Awesome page, though!

Is there a sort of "axiomatic high school maths" course, that builds up from the basics without handwaving? i.e. that covers the actual reasons and proofs for things, not just the mechanics/tricks/techniques. The way geometry is taught.

Some might be beyond high school level, like fundamental theorems of arithmetic (unique prime factorization) and algebra (polynomial factors/zeros). There's also non-integer exponents, defining reals etc.

All this stuff is used in High School but must be taken on faith. Good for believers, bad for skeptics.

Or is this just Mathematics 101? [my engineering maths didn't cover it]

Not a course but I believe

Mathematics: Its Content, Methods and Meaning by A.D. Aleksandrov, A.N. Kolmogorov , M.A. Lavrent'ev

may cover this in book form. (FWIW I have not read this, but intend to at some point)

It’s an amazing book! And very well written. Totally understandable!

I highly recommend the courses on brilliant.org. They are not axiomatic, but they heavily focus on building up deep insight rather than mechanical problem solving -- https://brilliant.org.

That said, for a rigorous proof-based approach to high school math, you may enjoy "Basic Mathematics" by Lang: https://www.amazon.com/Basic-Mathematics-Serge-Lang/dp/03879...

What about high-quality intros and tutorials you can complete in less time [than a serious academic course] to acquire practically reasonable understanding of particular technologies/subjects?

You know these are all affiliate links?

I'm curious - do you think affiliate links are a problem, and if so, why?

I have seen lots of udemy affiliate link spam on "learn programming" type subreddits. They use alt-accounts for fake reviews and make affiliate links looks like regular links. I'm okay with them as long as there is transparency and don't start a spam race.

they are affiliate links, yes.

The calculus course appears twice (same eventual URL), with different HN citations.

> Calculus: Single Variable Part 1 - Functions Coursera · University of Pennsylvania https://www.coursera.org/learn/single-variable-calculus

I took the one and only Linear Algebra course cited in this list and did not care for it. I found Strang's MIT course on the subject much better. I seem to recall reading other's had the same experience.

Perhaps we need a separate index that's just YT (and Vimeo?)?

The pop-up to subscribe every 2 seconds is extremely off-putting. Made me want to just close the tab multiple times. The curated content only kept me from doing it.

Edit: Nice work! Next steps would be to sort and collate similar topics.

Ick, that is a bug. Mind sending me details on your browser setup? yaj@the-domain-of-the-linked-story.

I like the idea. Would be interesting to see Gadgets recommended by HN users, Travel destinations recommended by HN users, Books recommended by HN users, Games recommended by HN users Movies recommended by HN users

As for books, there's already Hacker News Books https://hackernewsbooks.com/top-books-on-hacker-news

Indeed. Categories where there are authoritative urls are the low hanging fruit.

A previous related work of mine is TechYaks [0], which ranks software tech talks using (among other signals) HN recommends.

I have another one in the pipe...

[0]: https://techyaks.com/

Recommendations need to decay by date more.

Eg, the Johns Hopkins University "Data Science" course isn't worth doing (unless maybe it has been updated?), and note that most of the comments are pre-2017.

It's an R course, but it teaches pre-Tidyverse R.

Also it's not a great course in that the assignments aren't synced to the lectures - you need to know lots of stuff which haven't been touched on to complete the assignments at each stage. (Source: I did it in 2014 or 15)

I've been looking for something like this, thank you a lot! I wonder if something similar exists for HN book recommendations?

Just Google HN book threads.

Awesome work!

It'd be neat if there was a search feature that let us search by subject or keyword or any other course metadata.

I am curious to know how you built it, do you keep polling for the new a items on the HN? How do you figure out if it's a course? Do you use regexes for it or something more sophisticated / complecated?

Looks like the list counts complaints that a course does not exist as a recommendation, or why is the Crypto II course on coursere in the list? What was the original start date, September 2014 or 2013?

Any list like this one for finance / investing material? Thanks for the resource - will be put to good use. I’d love it if there were categories for paid, non-paid and overall most recommended.

Chalk me as one of those interested too. I decided that it was worth learning mastering that domain rather than continue on expanding my technical programming skills.

The biggest concern I have with this is that it promotes what's already popular to the top.

This means Blockchain and crypto are there and I doubt that's the most valuable thing you can learn this year.

I didn't even know there was an academy TLD. Is it saturated yet?

I’m planning to enroll in statistical mechanics course. Anyone else wants to team up and study as a group? We may finish faster that way. Email me: paras1987 <at> gmail

I'm looking for, and haven't yet found, a practical-as-possible, intermediate language design |& compiler design MOOC video course. Ideas appreciated.

I haven't taken these courses yet but Automata Theory is in my todo list. These are offered through Stanford Lagunita (Stanford's implementation of openedX)

* Compilers: https://lagunita.stanford.edu/courses/Engineering/Compilers/...

* Automata Theory: https://lagunita.stanford.edu/courses/course-v1:ComputerScie...

Recommendations and upvotes are a neat little signal to make sticky content rise to the top. This site is a great example.

I wonder if anyone ever made a list of paid only courses worth it. Now that would be really novel.

Hmm... where are Gilbert Strang his courses? He is recommended quite a bit, isn't he?

It would be helpful to be able to filter between free and paid off the main list.

This is a useful site, but you should probably disclose on it somewhere that you're getting affiliate revenue from every person clicking through and signing up for Coursera courses through your links...

I do, on every course page. Big grey box.

I missed that and I am an idiot. Mea culpa.

AWS, Udemy and others also offer a bunch of free coursers.

It's incredible how these free courses are rolling out.

Love this! Thank you!

Where's SICP ?

This is awesome!


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact