Profile cover photo
Profile photo
Jordan Tigani
Jordan Tigani's posts

Post has attachment
I'll be speaking about using Google Cloud Platform (particularly BigQuery) for IoT at RoboBusiness Europe in Denmark on June 2. They've put together a nice teaser video. Some of the other speakers are quite impressive. If you like big 'bots, and can make it to Odense in early June, check it out.

Post has attachment
My talk from the BigQuery meetup in Stockholm is now public. Mostly it is an introduction, but has some cool demos.

Post has attachment
Interview with BIME Analytics. I suppose a little bit of hyperbole never hurts.

Post has attachment
For anyone in the Seattle area that likes big data and cannot lie, Queen Anne Book Co is graciously hosting a launch party for our book. No purchase necessary, just come, have a glass of wine and a big data themed snack in a great local bookstore.

Since this is the first time +Siddartha Naidu and I have been in the same city since the book came out, we might have to do rap in SQL-92 to celebrate.

Post has attachment
I'm pretty excited about this post. Not because we've done so well on the predictions (more on that in a minute), but because we've been able to turn the predictions over to anyone who wants to give them a try.

When +Felipe Hoffa and I  first started working on the Google I/O talk that was the background for this effort, we did so because we wanted to demystify machine learning. Many developers and technologists think that ML is 'hard' and so don't think about all of the ways that it can work for them. However, between the open source tools that are available and Google's cloud, it is now pretty easy to do a lot of things that look 'hard'. We applied those tools to something that both of us are passionate about (soccer) and used them to make some predictions.

So now we've packaged up the models that we've built in a way that anyone can see exactly what we did and try them out themselves.  If you have an idea, it should be easy to incorporate in your model. There is lots of room for improvement.

It turns out that it is pretty easy to set this up -- the detailed instructions are in the post: just cut and paste a couple of command lines (you don't even need to modify them), then navigate to the iPython notebook in your web browser. That's all you need to do to start making predictions of your own.

And about that prediction accuracy. We've gone 13 for 14 so far, but I'd like to call out that this is a lot of luck. Soccer is just not predictable at that level of accuracy.  This world cup knockout stage was particularly surprising in that the favorites all won (so far). (See Nate Silver's take on it here:  So please don't think that these models are some magic oracle that will tell you who will win upwards of 90% of the time. At best, they'll tell you who should win. Or who would win more often if the game was played under the same conditions 100 times.

So give this a try and start predicting. I'd be happy to hear your results.

Post has attachment
Here are my predictions for the quarterfinals:
Brazil (71%) vs. Colombia (29%)
France (69%) vs. Germany (31%)
Netherlands (68%) vs. Costa Rica (32%)
Argentina (81%) vs. Belgium (19%)

Three of these are pretty uncontroversial... however picking France over Germany is probably the biggest surprise. This may be due to France having a softer route in the group phase than Germany, but also it looks like France has been more consistently solid than Germany. France was only scored on in one game, and that was in a 5-2 drubbing of Switzerland. They also won all of their games outright. Germany ended up drawing with Ghana and had the Algeria game been only 90 minutes long, that would have been a draw as well.

Lastly, I want to say that these predictions are just for fun; they were something that I worked on in spare time with the goal of demonstrating Machine Learning techniques for our I/O talk. I've been pleasantly surprised that the predictions were at least reasonable; that they did so well on the round of 16 was a lot a big dose of luck.

Post has shared content
Our talk on using machine learning to predict soccer outcomes managed to predict all of the round of 16 games correctly in the world cup.

This wasn't our goal; we just wanted to demonstrate different approaches to machine learning with Google's Cloud. But it is still nice to be right. 

I've been tweaking the models slightly and have predictions for the next round. However, this is where it starts to get a little bit since you don't have many obvious mismatches to get "for free". And our luck has got to run out sometime :-)
Machine learning and soccer: We predicted the results for 8 matches, we just got the 8 right! +Jordan Tigani and me at Google I/O 2014: Google I/O 2014 - Predicting the future with the Google Cloud Platform

Post has attachment
+Felipe Hoffa and I are going to be giving a talk at Google I/O on using machine learning with the Google Cloud Platform to make predictions, specifically to figure out who is going to win the World Cup. Check out the live stream (or attend in person, if you're going to I/O) if you're interested.

Post has shared content
Cool new dataset hosted in BigQuery.  +Ilya Grigorik ... any chance of adding this GDELT as a tag in My guess is that there are a lot of queries on this data that might interesting to a broad audience.

Post has attachment
My book "Google BigQuery Analytics" is now available on Google Play and wherever fine Electronic Books are sold (I hear that there is a company that is named after a south american river that also has it, but as the spouse of an independent bookseller, I cannot bring myself to link to it.)

The dead tree version should be available in a couple of weeks (June 9th, as far as I know) for anyone who needs a monitor stand. 
Wait while more posts are being loaded