Hacker News new | past | comments | ask | show | jobs | submit login
Data Science: CSCI E-109 Publication Listing (dce.harvard.edu)
284 points by tempw on Dec 4, 2016 | hide | past | web | favorite | 19 comments

The URL was changed from http://cs109.github.io/2015/pages/videos.html , which has more information and links to the corresponding repositories.

My project is at 28:00 of the final project presentations. We won overall that year. I'm TF (teaching fellow) for the current course and Head TF for the advanced course (CS109b) in the Spring. AMA.

For the lazy, it's chanceme.info, a data-driven college admission prediction project - that is really neat. Unfortunately, I'm getting a "Error: HTTP error 404." on the result.

Looks like an old DNS name expired. It has now been renewed.

Its really slick! I tried using it, but seem to be having the same issue :/

Bet a place like collegeconfidential.com would go bonkers over this tool :)

With only 13,000 training entries, I humbly think we built a better predictive model than anything else we could find. Certainly way better than Naviance. All the code and analysis is available on github if someone wants to make a business out of this.

Looks like a kitchen sink course. The deep learning lecture tried to summarize all of deep learning into 1 hour. Didn't do it for me.

This is how I prefer it. Give me the lay of the land, if you will, then I can dive into what is interesting.

There's nothing wrong with a kitchen sink course. Mostly likely the in-depth deep learning classes have their own course.

I'm taking this now, this year they split it into two classes, and so now there's much much more time to go into theory. But this is still an applications-focus class. They teach you general stuff: What to model, how to build, evaluate and improve models, how to do this on real world data, common pitfalls, how regression and classification works, then some specific models, tradeoffs and similarities between those models, then some methods to improve results from those models.

But this is great, because if you want to learn more in depth about those things, you can just take the second class in the series or a more advanced classes, and / or pick up a book, and you're already at a great position to improve.

I found the lecture useful. I haven't made it to the end, but there is a good visualization on solving the XOR problem without the Kernel Trick, starting at 26:57. She shows a multi-layer perceptron and uses two hyperplanes (hidden layer) to write a remapping so that a single hyperplane (output layer) can classify the 4 points. It is a general perceptron example, so it's only one hidden layer, but it shows the gist of DL (transforming data into a new representation) in a way that I hadn't seen before.

The only prereqs for this course are an introduction to programming course and an introduction to statistics course, so this isn't surprising.

Does the "1" in "109" imply that this is a intro course?

No, those would be two digit courses. 1xx courses are typically for either undergrad or grad credit. 2xx courses are typically for grad credit. CS50 is the classic entry level CompSci course at Harvard.


Do different Unis in the US use different numbering schemes?

Yes, while there are often some reused patterns (eg course 1xx being an intro course in many colleges) as we see here it certainly doesn't hold everywhere. For the most part there's no intercollegiate guarantee that bio150 at college X == bio150 at college Y.

Really great, would be cool if it were on youtube though

Some may find that useful but the video player used here (as with the one used for lectures at my university) supports far more relevant features such as browsing by slides and changing the layout of simultaneous video streams.

Like vinay427 said, I like the player here.

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