This competition is geared toward students passionate about machine learning and materials science to help unlock the power of open scientific data.
Participants will advance materials research by promoting open data and applying materials informatics to their data sets, all while collaborating with other students and earning prizes for their work.
Webinar 2: Tools and Workflows
Available January 17 at 2:00 pm ET
Eligibility and Prizes
Machine learning and data-driven techniques are playing a big role in the research, development and discovery of new materials. These approaches to materials R&D can reduce the time to reach conclusions in research, reduce computational costs and contribute to significant cost savings for materials researchers and manufacturers.
The backbone of materials informatics and machine learning is data, but much of it is locked away in lab notebooks, research papers, or on personal computers.
The purpose of this competition is for teams of students to curate or develop their own materials data set, and apply materials informatics to that data set to learn something new, whether it’s previously unknown correlations in process–structure–property linkages, a promising new material that meets a desired performance target, or a machine learning model that approaches experimental or computational results.
Compete for cash prizes ($US):
- 1st place – $1000
- 2nd place – $500
- 3rd place – $250
Winners will be announced during the 2019 MRS Spring Meeting in Phoenix, Arizona, and have their poster included in the Spring Meeting and/or an MRS Bulletin article.
- Create or curate an open materials data set from a literature review or from your research groups’ experimental or computational data. Alternatively, you can present a novel analysis of a previously available open materials data set.
- Label the data set and data points with sufficient metadata and citations so someone else can understand how the data were generated.
- Analyze your data using materials informatics.
- Describe your methodology—how did you analyze your data? What tools did you use? What did you learn? If someone else wanted to repeat this methodology, describe the steps they would take to implement your analysis, including links to the tools you used.
- Create a poster or visual presentation that summarizes your work. Submissions must be in one of the following formats: 10 slides (.pptx, Prezi, Google slides, .odt), a PDF research poster, iPython notebook, or web applet. Please review your submission and ensure that it displays properly. Your presentation should include:
- Objectives—Why did you build this data set? What did you hope to learn with your analysis?
- Analysis and Results
The purpose of this competition is to promote open materials data and materials informatics techniques. Using an open-source data repository is required, and using open-source materials informatics tools is encouraged.
Please indicate your interest in the competition via this form. We will follow up with additional details and helpful resources to get you started.
Student groups will be judged by materials scientists, data scientists and engineers on the novelty of their open data set, the clarity of their methodology, the quality of their presentation and the scientific value of their analysis.
Contact Joshua Tappan (firstname.lastname@example.org
), Community Manager at Citrine Informatics, with questions.