probabilityOfClosing = (question) ->
text = question.text.toLowerCase()
return (text.length / (text.indexOf('jquery') + 2)) / 100Stack Exchange Machine Learning Contest
21–30 of 67 posts
Re: Stack Exchange Machine Learning Contest
#22If you're someone who's interested in ML/datamining but haven't had a chance to put your ideas to any hard/interesting problems I strongly recommend a kaggle contest. It's one thing to plug some data into a random forest and go "oh cool, I guess that did okay" and entirely another to see how other competitors are comparing. one of the biggest challenges I've found in implementing ML projects is I don't have a great s…
This is a great idea! I just signed up. So let me ask you, where do you read up on how past winners have won? How did you decide on algorithms to try out on a contest? How did you find promising academic papers?
Kaggle has blog posts of many of the past winners: http://blog.kaggle.com/category/dojo/
For algorithms, just try whatever you know best/is fastest to implement. If you're using R I highly recommend the Caret package.
For papers: the best place to get started is to begin browsing the forums or any similar contests, the community there is actually pretty awesome and will frequently post papers. After that google scholar (or even just google) for particular problems will yield nice results.
Also checkout the wiki: http://www.kaggle.com/wiki/Home
Re: Stack Exchange Machine Learning Contest
#23If you're someone who's interested in ML/datamining but haven't had a chance to put your ideas to any hard/interesting problems I strongly recommend a kaggle contest. It's one thing to plug some data into a random forest and go "oh cool, I guess that did okay" and entirely another to see how other competitors are comparing. one of the biggest challenges I've found in implementing ML projects is I don't have a great s…
I am really interesting in ML, but have only recently been diving into it. I have watched all of the videos for Andrew Ng's Coursera course (and most of the programming exercises), but just looking over some of the Kaggle contests I think I would be quickly out of my depth. Would I be wasting my time attempting these with such a basic level of knowledge?
If you read through the past winners you'll find that in many cases a very simple model will win. I believe one of the winners that posted a blog post had pretty much the background you describe.
When I started I was in a similar position to you and just wanted to see if I could even tread water with some of the really knowledgeable members of the community. I ended up placing in the top 5 for one of the contests I was in (with btw a really simple model).
They usually give you some starter code in either R or Python which will give you the results for a benchmark, start there and then use cross-validaton to see if you can beat that bench mark, and if you do submit. It's very addictive and you'll come away knowing a lot more than you started with.
Re: Stack Exchange Machine Learning Contest
#24It's nice to see this kind of contest but the topic just sets me off on a much-needed rant. The moderator situation on Stackoverflow is getting out of control. I see a Q&A site as having three main groups: 1. People who ask questions; 2. People who answer questions; and 3. People who edit/moderate questions. Even 2+ years ago there was a lot of lip service paid to the value of (3). I disagreed then and it's only been…
How do I contact you? The Kontactr form at http://www.cforcoding.com/ is broken..
Re: Stack Exchange Machine Learning Contest
#25Earlier quoted context omitted.
Does every Stack Exchange post need to be hijacked by yet another whine about site policy and moderation activities?
This particular post is about replacing human moderators with bots, whining about SE moderation policy is actually quite related to the post.
Replacing moderators with bots would be a terrible idea and hopefully will never happen. Kevin Montrose clarified it in a comment:
> To be clear, we’re not intending to have an algorithm automatically close posts. This is aimed at improving our quality metrics (which guide users before posting) and auto-flagging facilities (which help focus moderation attention).
Re: Stack Exchange Machine Learning Contest
#26Earlier quoted context omitted.
> But consider a question like "What are the pros and cons of Sinatra vs Rails?" This sort of question (IMHO) absolutely has value as someone experienced with both could enumerate the relative merits of each in a pretty objective fashion without making an absolute determination. This is something that absolutely could have value to anyone evaluating Ruby Web frameworks. I guess, but Zookeepers could also potentially…
> I guess, but Zookeepers could also potentially talk about "What are the pros and cons of Gorillas vs Sharks?" And on an appropriate forum with as many readers as SO/SE you'd likely get well-thought out responses going far into detail regarding purchase cost, habitat maintenance, prevalence of skilled keepers, etc. Someone who actually was deciding whether to add a Gorilla enclosure or a shark aquarium would find it…
You hit the nail on the head... "an appropriate forum with as many readers as SO/SE"
Stack overflow on the other hand is for people to:
[A]sk practical, answerable questions based on actual problems that you face. Chatty, open-ended questions diminish the usefulness of our site and push other questions off the front page.
Re: Stack Exchange Machine Learning Contest
#27Earlier quoted context omitted.
No, but it's particularly apposite on this post, as SE is asking for a machine learning algorithm is to assist in moderating questions: an algorithm that predicts whether (and for what reason) a question will be closed. which raises all sorts of questions as to what the reasons are for which questions [should][1] be closed - this is a grey area, and there is much argument on SE itself over what sort of questions are…
Worth a try in the name of science, but I'm pessimistic about the ability of machine learning to measure subjective "quality" of a question. In the best case you'll end up with a txtspeak and Proper English detector. But I knew that already.
I'd be interested to see if there's any predictors for "good quality" questions vs anything else. Some research project for students, perhaps.
I remember reading something about telling the difference between truthful and fake reviews.
(http://cornellsun.com/section/news/content/2011/09/23/cornel...)
There's more literature in this paper.
(http://www.cs.uic.edu/~liub/publications/WWW-2012-group-spam...)
Re: Stack Exchange Machine Learning Contest
#28It's nice to see this kind of contest but the topic just sets me off on a much-needed rant. The moderator situation on Stackoverflow is getting out of control. I see a Q&A site as having three main groups: 1. People who ask questions; 2. People who answer questions; and 3. People who edit/moderate questions. Even 2+ years ago there was a lot of lip service paid to the value of (3). I disagreed then and it's only been…
"Take this question from Stackoverflow: Database development mistakes made by application developers [3], a question I spent some time answering and that people seemed to appreciate the answer to (based on comments and 1000+ upvotes). It is closed as "not constructive". This is hardly a unique phenomenon. We've all seen many interesting questions posted here that are now closed or locked and who knows how many have been deleted."
This is a rant that's needed and one that SO needs to open themselves to receiving. Their stance on this is just wrong.
I'd actually bookmarked several of your posts (among others) because they were so valuable. So it enrages me to no end to click those bookmarks and find that the entire discussion is simply gone.
To recap - a high quality contribution whose value was validated by dozens of individuals (or perhaps even more) was simply deleted.
The problem starts at the top. As you can see from Jeff Atwood's post below (codinghorror) , even the founders of Stack Overflow don't understand the value of their own platform to their customers. They had a preconceived notion of what StackOverflow Is and they are going to stick to it, users be damned.
I don't mean the bash them, but if nerds want to see a prime example of why they take orders from non-nerds (whom we like to think of as "less intelligent" than us), this is exhibit "A". And if you want to know why StackExchange will fail everywhere else, this is exhibit "A". No other group of users will put up with that crap.
Re: Stack Exchange Machine Learning Contest
#29Earlier quoted context omitted.
What are common database development mistakes made by application developers? It is a question, an open question intended to provoke debate and teach about a subject, in effect it's a request for an FAQ. Now perhaps SO is not intended to be for that sort of question, and that is of course for SO to decide. I suppose the reason many people come to SO to read questions is that they'd like to learn about a subject area,…
> or maybe SO should just expand to encompass that sort of FAQ function? Some of the tag wikis do this: http://stackoverflow.com/tags/java/info http://stackoverflow.com/tags/c%23/info So Cletus' post might make more sense as the tag wiki for, say: http://stackoverflow.com/tags/database/info
Yes it could although the wikis don't contain the other alternative questions or the dialogue which make Q&A sessions involving several people who know the subject well useful for other readers. For many questions (particularly open ones) there are many answers, and no right answer for all circumstances, so there is no clear division between this question/answers and other longer more specific ones as to being opinion or fact - they're mostly a mix of both.
Because people will inevitably continually ask/answer these questions and many similar more specific and yet still open ones and see the resulting debate as useful on any QA site, it might be good to have a more structured way of moving them to an FAQ section on SO, without destroying all the ad-hoc relations and rewards that users have built up using your QA format (i.e. not turning them into a wiki, which doesn't really suit them and loses all the attribution, comments, discussion etc).
It feels a little draconian at present sometimes when useful answers are marked as 'trivial' or 'not constructive' when they clearly are constructive, but are constructive in a direction SO didn't anticipate.
Re: Stack Exchange Machine Learning Contest
#30Earlier quoted context omitted.
I am really interesting in ML, but have only recently been diving into it. I have watched all of the videos for Andrew Ng's Coursera course (and most of the programming exercises), but just looking over some of the Kaggle contests I think I would be quickly out of my depth. Would I be wasting my time attempting these with such a basic level of knowledge?
If you read through the past winners you'll find that in many cases a very simple model will win. I believe one of the winners that posted a blog post had pretty much the background you describe. When I started I was in a similar position to you and just wanted to see if I could even tread water with some of the really knowledgeable members of the community. I ended up placing in the top 5 for one of the contests I w…
Awesome, thanks for the info. I am checking out some of the benchmarks now.
Why do you think it is that simple models often win? Is it due to the experts no participating or is there a lot more low-hanging fruit than I previously thought? Or just that simple models are easier to use and reason with for humans and thus easier to get right.