# DataScience Central competition: Automate jackknife regression

Data Science Central holds a competition to get statisticians more involved in Data Science - create a black-box, automated, easy-to-interpret, sample-based, robust technique called jackknife regression.

By Vincent Granville, DSC, Apr 1, 2014.

I'd like to get statisticians more involved in data science, and here's an opportunity to participate, earn money, and make statistical science more visible.

Data Science Central (the leading community for analytic practitioners with 3 million visitors per year and more than 200,000 members) organizes a competition to write a research paper about a statistical technique called jackknife regression (a robust regression technique).

In connect with our proposed methodology, the goal is to create a black-box, automated, easy-to-interpret, sample-based, robust technique called jackknife regression, to be used in small and big data environments by non-statisticians.

The details are found at bit.ly/1flGVks (on datasciencecentral.com). It involves working on simulated data. The results, to be published on our network and possibly in a scientific journal, will reach far more practitioners than any article published in a statistical journal. The award is $1,000.

On a different note, if you want to get your students interested in experimental design (or criticize flaws in my 'data science' version), you feel free to share my interactive, real-time experiment. You can find it, participate and check results in real-time at bit.ly/1oqj1XB (also on datasciencecentral.com).

I'd like to get statisticians more involved in data science, and here's an opportunity to participate, earn money, and make statistical science more visible.

Data Science Central (the leading community for analytic practitioners with 3 million visitors per year and more than 200,000 members) organizes a competition to write a research paper about a statistical technique called jackknife regression (a robust regression technique).

In connect with our proposed methodology, the goal is to create a black-box, automated, easy-to-interpret, sample-based, robust technique called jackknife regression, to be used in small and big data environments by non-statisticians.

The details are found at bit.ly/1flGVks (on datasciencecentral.com). It involves working on simulated data. The results, to be published on our network and possibly in a scientific journal, will reach far more practitioners than any article published in a statistical journal. The award is $1,000.

On a different note, if you want to get your students interested in experimental design (or criticize flaws in my 'data science' version), you feel free to share my interactive, real-time experiment. You can find it, participate and check results in real-time at bit.ly/1oqj1XB (also on datasciencecentral.com).