This has already been posted to Hackernews! Having said, I used Prediction.io all day today, (against a few years of Shopify Ecommerce order data) and it is the easiest way to break into machine learning. Props to the Predictionio team!
Just being curious, what kind of metrics are you interested in predicting for Shopify? I would give away my car in exchange for a few years of Shopify order data :)
I've received similar email about possible integration with Pyramid framework last week. So I created sample package [1] but haven't got any reply yet.. Hate when people don't respond to emails.
[1] https://github.com/hanula/pyramid_predictionio
I got an email like grandparent and I replied asking about licensing. They replied, but it took a a little over 2 weeks before they replied. So, they will probably reply, just not promptly.
While I'm not a huge fan of people spamming email address they grabbed from Github (Yes I know I can hide my email) I have been contacted because of my GitHub account (because of activity/repos/etc) in the past and had no problem with it. I just don't like mindless spamming which this appears to be. That said where do you draw the line (and do you think this email crosses said line)? (I'm genuinely curious)
It's not illegal in the US unless they don't provide opt-out and the various other things CAN-SPAM demands. That said, starring a repo could potentially be considered forming a relationship and relationship messages are exempt from those rules.
The project looks good, however I got this same e-mail, I was planning to test the PHP-SDK and the Python-SDK and test the product, however now I see, I was just the target of a github SPAM. =/
Same, asking about suggestions in terms of best practice to integrate PredictionIO in some PHP frameworks.
If I like the product, I may open-source some plugin or wrapper around their SDK for specifics product or framework, but asking will not make it happens faster.
I also got an unsolicited email from them asking for feedback on the product. I didn't really mind since it is relevant to my area, but it was still a bit odd.
The MOOC course is very interesting. Do you know if there's a way to get enrolled? It looks like the registration is closed and there's no upcoming offerings available currently. Thanks.
Since the authors claim that it is "Built on top of scalable frameworks such as Hadoop and Cascading" I guess that it is aiming to be production-ready while scikit is mostly for prototyping.
In that case then what is the added value compared to Mahout, the Hadoop-based ML framework?
I'm always extremely skeptical of such initiatives, because ML is not a magical black box where you put your data in one end and you get results on the other end. Automating the trivial parts of ML, ie. providing an API to a ML library, is a week-end project, but that in itself is useless. If you don't automate the hard parts of ML, such as feature engineering, then you're not providing any value at all.
Scikit-learn is a library of machine learning algorithms.
PredicitionIO does the stuff around machine learning you need to get a system into production. It provides a REST API to interact with the system and it handles data storage and retrieval as well. I think it has fewer algorithms than scikit-learn. This may or may not be a problem depending on your use case.
I'm pretty sure it's just really well recorded. The voice changes in tone to emphasize different words, which isn't a feature in any text-to-speech program I know of.
All that said, every time I read a comment in this thread, I hear it in that voice now.
Also got an email, because I'm "engaged" with Django framework on Github but I hadn't starred PredictionIO at this point. Sent a reply, no response. Seems they just want visibility among devs and are using the fact they're open-source as an excuse to directly contact people. Good product or not I really don't like being deceived.
> r = cli.get_itemrec_topn("myEngine", 5, {"pio_latlng":[37.9, 91.2]})
Without the preceding comment I'm sure this line of code is clear as crystal to whomever wrote it. To me it might as well be hieroglyphics. If you write an API, do not abbreviate the public methods.
Totally unrelated, but looking in the source for PredictionIO (which is Scala) they've got this really deep directory nesting before you can find the code. That's something that always annoyed me about typical Java apps.. is that also a common thing in the Scala world?
If you use the sbt build tool, which is by far the most popular in the Scala world, you have to use Maven-style directory layout which is the originator of the horror you refer to (see http://www.scala-sbt.org/release/docs/Getting-Started/Direct...). Most Scala projects don't bother with the com/foo/mystuff directory structure part, retaining just the mystuff directory, which is a small saving.
It's an annoying vestige, but I typically don't navigate the directory structure to find files anyway. Instead I use projectile in Emacs. I'm sure other editors have similar systems.
Recently exposed to http://0xdata.com/ at Strata NYC and was very impressed. Just another to add to the list of great open source modeling and prediction tools.
I also received a spam email from them after contributing to Rails and sent them an email about it. PredictionIO people: your fellow developers aren't idiots. You wouldn't want to be spammed after contributing to open source projects, so don't spam us.
46 comments
[ 5.0 ms ] story [ 73.3 ms ] threadWhile I'm not a huge fan of people spamming email address they grabbed from Github (Yes I know I can hide my email) I have been contacted because of my GitHub account (because of activity/repos/etc) in the past and had no problem with it. I just don't like mindless spamming which this appears to be. That said where do you draw the line (and do you think this email crosses said line)? (I'm genuinely curious)
[1] - http://www.campaignmonitor.com/guides/permission/spam/
If I like the product, I may open-source some plugin or wrapper around their SDK for specifics product or framework, but asking will not make it happens faster.
and vowpal wabbit https://github.com/JohnLangford/vowpal_wabbit/wiki
[0] - http://www.cs.waikato.ac.nz/ml/weka/
I'm always extremely skeptical of such initiatives, because ML is not a magical black box where you put your data in one end and you get results on the other end. Automating the trivial parts of ML, ie. providing an API to a ML library, is a week-end project, but that in itself is useless. If you don't automate the hard parts of ML, such as feature engineering, then you're not providing any value at all.
All that said, every time I read a comment in this thread, I hear it in that voice now.
http://myrrix.com/
Without the preceding comment I'm sure this line of code is clear as crystal to whomever wrote it. To me it might as well be hieroglyphics. If you write an API, do not abbreviate the public methods.
It's an annoying vestige, but I typically don't navigate the directory structure to find files anyway. Instead I use projectile in Emacs. I'm sure other editors have similar systems.