Skip directly to content

'Divvy' app opens up machine learning

NSF Award:

Divvy: Robust and Interactive Cluster Analysis  (University of California-San Diego)

State:
Congressional Districts:
Research Areas:

The field of machine learning has produced many valuable data analysis techniques. However, sharing these techniques with users outside the field is challenging because of their complexity. New algorithms are often difficult to learn and apply correctly. For example, the authors may fail to provide an example implementation of the algorithm, or may supply code for an unfamiliar language or platform. Additionally, new algorithms might make strong assumptions about the structure of their inputs, or be very sensitive to parameter changes.

Now a free, open-source app called "Divvy" makes sophisticated machine learning algorithms more accessible to a wider range of researchers. Developed by researchers at the University of California, San Diego, Divvy provides a platform for machine learning researchers to publish fast, easy-to-use versions of their algorithms. With Divvy, researchers can quickly run an assortment of algorithms on their data without worrying about programming languages, data formats or visualization.

Images (1 of )

  • divvy app groups images of the same object in a 2-d scatter plot
  • divvy users can run and display multiple analyses simultaneously
Divvy groups images of the same object in a 2-D scatter plot.
Joshua Lewis and Virginia de Sa, University of Rochester
Divvy simultaneously runs and displays multiple analyses.
Joshua Lewis and Virginia de Sa, University of Rochester

Recent Award Highlights

image and schematic show how the human brain analyzes biological motion

Distinguishing between animate and inanimate objects

Animacy judgments are governed by physical and biological constraints

Research Areas: People & Society Locations: California
regional culture and a major university are keys to successful innovation economies

The role of innovation in regional economies

Regional culture and a major university contribute to innovation economy success

Research Areas: People & Society Locations: California