Now live: Innocentive Marketplace. Get shortlisted for funded opportunities today.

Efficiently Recognize Anomalous Data in a Database Regardless of Context

17 Submissions
$30,000 USD
Challenge closed

Challenge overview

The Seeker desires a theoretical method to efficiently find rare errors in a database, without relying on manual case-by-case rules or context-sensitive parameters. 

This Challenge requires only a written proposal.

The Seeker conducts clinical trials and manages the resulting information in large databases.  As numerous investigators collect information, the investigation grows in complexity and scale, and it becomes increasingly difficult to ensure that each data point accurately describes the situation under investigation.  Because the quality of the data in the database determines the quality of conclusions that can be drawn, quality control is very important. 

Straightforward “first pass” data quality filters are routinely implemented to ensure syntactic and logical compliance.  However, it is more difficult to identify deeper errors.  Thus, the Seeker desires a methodology, framework, or approach to identify rare errors in a large database, without resorting to extensive case-by-case manual rules.  The Solution should require minimal human intervention to accurately detect errors. 

This is a Theoretical Challenge that requires only a written proposal to be submitted.  The Challenge award will be contingent upon theoretical evaluation of the proposal by the Seeker. 

To receive an award, winning Solvers will have to transfer to the Seeker their exclusive Intellectual Property (IP) rights to the solution.

Use the slider to explore how the Challenge process works:

Register

Review & Accept

Submit

Win

To start solving this Challenge, log in to the Challenge Center or register as a Solver