FAQ

  1. Can several people enter as a team?
    Yes, a team may participate. The only requirement is that each person on the team must register and click the "Download data" link in the participant's dashboard to sign the non-disclosure agreement (NDA). Please make note of all the team members on your submission.
  2. I am overseas from the United States. Since the event will be held in the USA, will I be able to participate?
    Yes, you are eligible to compete in the contest. If you are selected as a finalist you will have the option to present via videoconference.
  3. Am I allowed to publish my entry in a journal?
    The work may be published per the Data Use Agreement you sign when downloading data for the competition.
  4. May I use the Syngenta AI Challenge data for educational purposes?
    Please refer to the Data Use Agreement for the Syngenta AI Challenge.
  5. I work for a large agricultural company, can I compete?
    Employees or people who are associated with large agricultural companies are not eligible to compete. Please contact us if you have any questions about whether or not you are allowed to participate.
  6. How do you handle spatial and temporal variations?
    In general varieties are selected in a year to basis. Results of field trials as the ones in the challenge are analyzed to estimate expected yield using analysis of variances. In them sources of variation are accounted for.
  7. Does elite = high yield or stable yields from year to year or location to location?
    Elite varieties should have yields across year and across locations.
  8. When we are evaluating best-performing varieties, is RM an important factor apart from Yield? In other words, when deciding whether a variety is 'truly' elite, is Yield the only factor that plays a role here? Or should we set up some gold standard to decide whether a variety is elite or not, i.e. considering RM as well?
    Yes it is. RM indicates the length of the life cycle. In General the longer the cycle the greater the yield… So I would recommend to group the varieties by RM band and rank them by performance.
  9. Yield is a complex trait - there is not one SNP that can predict it. How should we use the SNP data?
    In any way you consider appropriate. Also you can choose not to use it or follow the literature.
  10. According to the “RM” factor’s description, it is implying that 0.1 RM = 1 day. Can we say 2.5 RM means it takes 25 days for a Variety to reach maturity?
    For a given location a variety of RM 0 would have a life cycle than a variety of RM 2.5.
  11. In the illustration below, 15% of varieties are selected from each stage to the next one. However, it seems that in the actual stage dataset we received, only 5% varieties in Stage 1 advanced to Stage 2. Can you help explain the selection process?
    The diagram is for illustration purposes. The range the percentage of varieties selected to advance in stage is variable. Than can also be observed in the dataset.
  12. Is there a unique link between varieties and genetic data?
    Yes, it is the Variety variable.
  13. In the sample submission you provided, you gave three files. Are we supposed to submit these 3 files for each of the three stages, or are we only required to submit the predictions for Year 2015-2016?
    You are only required to submit the predictions from 2015-2016.
  14. We found it hard to make predictions for new varieties in Class of 2014, which didn't exist in previous years. We can only know about their geographic and temporal information but not anything about their intrinsic variety characteristics. We wanted to infer their variety characteristics from their genetic info; however, the genetic info file we have now doesn't provide data for all the varieties.
    Could you provide us with any advice on dealing with those novel varieties? Or if possible, will you be able to provide us with the complete genetic information for all the varieties that is in the data set?

    As with many real-world datasets, there exist gaps in the seed data. However, Stage III data (released May 1), which contains data describing varieties of the 2014 class that cover all experimental trails for year 2014, should be able to help.