Web-based system automatically evaluates proposals from far-flung data scientists

October 30, 2017 by Larry Hardesty, Massachusetts Institute of Technology
“I think that the concept of massive and open data science can be really leveraged for areas where there’s a strong social impact but not necessarily a single profit-making or government organization that is coordinating responses,” MIT graduate student Micah Smith says about FeatureHub. Credit: Massachusetts Institute of Technology

In the analysis of big data sets, the first step is usually the identification of "features"—data points with particular predictive power or analytic utility. Choosing features usually requires some human intuition. For instance, a sales database might contain revenues and date ranges, but it might take a human to recognize that average revenues—revenues divided by the sizes of the ranges—is the really useful metric.

MIT researchers have developed a new collaboration tool, dubbed FeatureHub, intended to make feature identification more efficient and effective. With FeatureHub, data scientists and experts on particular topics could log on to a central site and spend an hour or two reviewing a problem and proposing features. Software then tests myriad combinations of features against target data, to determine which are most useful for a given predictive task.

In tests, the researchers recruited 32 analysts with experience, who spent five hours each with the system, familiarizing themselves with it and using it to propose candidate features for each of two data-science problems.

The predictive models produced by the system were tested against those submitted to a data-science competition called Kaggle. The Kaggle entries had been scored on a 100-point scale, and the FeatureHub models were within three and five points of the winning entries for the two problems.

But where the top-scoring entries were the result of weeks or even months of work, the FeatureHub entries were produced in a matter of days. And while 32 collaborators on a single data science project is a lot by today's standards, Micah Smith, an MIT graduate student in electrical engineering and computer science who helped lead the project, has much larger ambitions.

FeatureHub—like its name—was inspired by GitHub, an online repository of open-source programming projects, some of which have drawn thousands of contributors. Smith hopes that FeatureHub might someday attain a similar scale.

"I do hope that we can facilitate having thousands of people working on a single solution for predicting where traffic accidents are most likely to strike in New York City or predicting which patients in a hospital are most likely to require some medical intervention," he says. "I think that the concept of massive and open data science can be really leveraged for areas where there's a strong social impact but not necessarily a single profit-making or government organization that is coordinating responses."

Smith and his colleagues presented a paper describing FeatureHub at the IEEE International Conference on Data Science and Advanced Analytics. His coauthors on the paper are his thesis advisor, Kalyan Veeramachaneni, a principal research scientist at MIT's Laboratory for Information and Decision Systems, and Roy Wedge, who began working with Veeramachaneni's group as an MIT undergraduate and is now a software engineer at Feature Labs, a data science company based on the group's work.

FeatureHub's user interface is built on top of a common data-analysis software suite called the Jupyter Notebook, and the evaluation of feature sets is performed by standard machine-learning software packages. Features must be written in the Python programming language, but their design has to follow a template that intentionally keeps the syntax simple. A typical feature might require between five and 10 lines of code.

The MIT researchers wrote code that mediates between the other software packages and manages data, pooling features submitted by many different users and tracking those collections of features that perform best on particular data analysis tasks.

In the past, Veeramachaneni's group has developed software that automatically generates features by inferring relationships between data from the manner in which they're organized. When that organizational information is missing, however, the approach is less effective.

Still, Smith imagines, automatic feature synthesis could be used in conjunction with FeatureHub, getting projects started before volunteers have begun to contribute to them, saving the grunt work of enumerating the obvious features, and augmenting the best-performing sets of contributed by humans.

Explore further: System that replaces human intuition with algorithms outperforms human teams

More information: Paper: "FeatureHub: Towards collaborative data science" dai.lids.mit.edu/featurehub-smith.pdf

Related Stories

Bug-repair system learns from example

September 29, 2017

Anyone who's downloaded an update to a computer program or phone app knows that most commercial software has bugs and security holes that require regular "patching."

Recommended for you

1 in 3 Michigan workers tested opened fake 'phishing' email

March 16, 2018

Michigan auditors who conducted a fake "phishing" attack on 5,000 randomly selected state employees said Friday that nearly one-third opened the email, a quarter clicked on the link and almost one-fifth entered their user ...

Origami-inspired self-locking foldable robotic arm

March 15, 2018

A research team of Seoul National University led by Professor Kyu-Jin Cho has developed an origami-inspired robotic arm that is foldable, self-assembling and also highly-rigid. (The researchers include Suk-Jun Kim, Dae-Young ...

Tokyo Tech's six-legged robots get closer to nature

March 12, 2018

A study led by researchers at Tokyo Institute of Technology (Tokyo Tech) has uncovered new ways of driving multi-legged robots by means of a two-level controller. The proposed controller uses a network of so-called non-linear ...


Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.