BigML News, Issue #4, September 2013
BigML - Machine learning made easy


Hi <<Full Name>>,

In this issue:
  • Back to School with BigML
  • Latest Feature:  Node.js bindings
  • Featured Model:  Etsy Store Success
  • Feature Highlight:  Evaluations
  • Machine Learning Explained: Mater over Mind in Machine Learning
  • New BigML Team Member:  David Gerster, VP of Data Science
  • What's Cooking in BigML's labs?
  • BigML User Survey—fill it out to get a free subscription!
Back to School with BigML
Back to school with BigML

While you've hopefully enjoyed a relaxing summer, we here at BigML have been hard at work—adding new features, growing the team, and establishing partnerships to help you be more successful with your predictive modeling activities.  The introduction of subscription plans gives you the ability to perform unlimited tasks with your data at 64MB, 1GB and 4GB levels, starting as low as $30/mo and even cheaper for quarterly, yearly or student/non-profit usage.

We're super excited that there are now over 5,000 BigML users, and you've collectively performed around half a million predictive modeling tasks.  Please be sure to fill out our semi-annual survey in order to provide feedback on BigML (and to receive a free one-month subscription!), and never hesitate to email us at should you have any questions or requests.

In support of the new school year, BigML is pleased to offer students 50% off of paid subscriptions—just remember to check the “student, NGO or non-profit” box before making your purchase.  Also, we're happy to support any classes and instructional programs on Machine Learning or Predictive Analytics. Contact us at to discuss discounted usage for class participants, collaboration on class materials and/or guest presentation.

Latest Feature: Node.js Bindings
Node.js Bindings

Node.js is a software platform that is used to build scalable network and server-side applications—we're excited to bring the power of Machine Learning to the node.js community through the release of our node.js bindings.  This allows node.js developers to create low-latency, event-driven applications that use our Machine Learning back end to add predictive attributes and capabilities to their projectwhich will be especially powerful when coupled with our forthcoming high-performance prediction service!

You can download the library from github, learn more from our documentation, or email us with any questions.

Featured Model: Etsy Success Predictions
Model of the month

BigML worked with's ( web crawler browser to pull data from nearly 60,000 Etsy stores into a single dataset, which we then imported into BigML to help predict likely sales from an Etsy store based on variables such as number of items offered, number of admirers, year of store establishment and amount of customer feedback.

You can view the model to see the results and start building your own predictions.  Or learn more about the work BigML did with in our blog post.

Feature Highlight: Evaluations
Feature Highlight: Evaluations

We often get the question “is it safe to predict with my model?”  We already answered this here. The best way to estimate your model's strength is to run an Evaluation.  This allows you to create a test set from your data to assess a number of measures like accuracy, precision, and recall in classification tasks or expected error in regression tasks. We have been working on making evaluations easier. Now you can create your training and test set splits in only one click. We have also added a nice set of alerts to tell when you cannot trust your model at all as it will perform worse than a model that just makes random decisions.

Machine Learning Explained: Matter over Mind in Machine Learning
Machine Learning Explained

Machine learning papers typically proceed as follows: find a dataset for a problem that is only partially solved, invent or improve an algorithm to solve this problem such that some improvement on some standard metric is made on that dataset, write it up and publish. Repeat until tenured. While the above approach produces a substantial number of computer science professors, it isn't nearly as successful at getting machine learning algorithms in a position to make a measurable difference in society at large. However, Dr. Kiri Wagstaff sees things very differently. She published a very inspirational paper at the 2012 ICML conference that laid out a program of challenges for machine learning researchers interested in making a difference in the real world. Inspired by that paper, our machine learning guru, Dr. Charlie Parker, wrote this great blog post.

New BigML Team Member: David Gester, VP of Data Science
New BigML Team Member

BigML is very pleased to welcome David Gerster to the team.  As BigML's VP of Data Science, David engages with a variety of customers and partners to ensure that BigML's product offerings are fully representative of customer requirements.

Prior to joining BigML, David was Director of Data Science at Groupon.  While at Groupon, David used BigML to help predict usage of Groupon's mobile application.  As David likes to tell people, he liked BigML so much that he decided to join us!

What's cooking in BigML's Labs?
What's cooking in BigML's Labs?

The BigML team is hard at work on some killer new features and functionality which we plan on rolling out to you over the next few months.  Keep your eyes and ears peeled for support for text processing, workgroups, high-performance predictions, time-series analysis, and added predictive model visualizations. 

BigML User Survey: Get a free BigML Subscription

We want your feedback!  Please complete our BigML user survey to let us know what you think of BigML, and to make recommendations on what we can do to better serve you.  Everyone who completes the survey will receive a one-month Standard Subscription (or, for current subscribers, up to $30 credit off their next month's invoice).

Our mailing address is:
BigML, Inc
2851 NW 9th Street
Suite D, Conifer Plaza Building
Corvallis, Oregon 97330

Add us to your address book

Copyright © 2013 BigML, Inc, All rights reserved.