Summarize Challenge

Teams:

• Plant Prodigies - Alexis and Michael

• Pickle Rick – Hongyuan and Ian

• Team X - Jake Newton

• Data Dudes- Raul Manquero-Ochoa

• Team Miracle Grow - Derek Colombo

• Cyber Crop-Bots - Nik and Abishek

Summary:

• Develop machine learning algorithms that can accurately predict yearly crop yield for a particular performance record.

• Use test data and validation data to ensure accurate predictions and be general enough to not overfit the data set (k-means).

• Edits and collaboration done through GitHub.

• Document procedures and processes well to ensure reproducibility and improve understanding across the competition.

• Template provided for the challenge: https://mlcas2021.github.io/files/MLCAS2021-template.docx

• Data set provided for the challenge: https://drive.google.com/file/d/1DoyextA0q4mxumMAhBvqZbfZriIM9A-Y/view

• Poster Template is yet to be released, but the dimensions are to be width around 84.1 cm and height around 118.9 cm. 

• Submissions through MLCAS via Microsoft CMT account: https://cmt3.research.microsoft.com/User/Login?ReturnUrl=%2FMLCAS2021

• The leaderboard ranks teams ML models using performance metrics such as RMSE (Root Mean Square Error), MAE (Mean Absolute Error), and R2_Score (coefficient of determination)

Evaluation Criteria

For evaluation, you will utilize the test dataset (10,337 samples) to predict the yearly yield for each sample. You are required to upload a NumPy file of shape (10337, ) comprising the test set yield predictions. The predictions need to be on the same scale as the yield values of the training dataset. 

"We will use Root Mean Square Error (RMSE) as the evaluation metric"

Important dates:

• Oct 18: Final Submission Deadline

• Oct 25: Announcement of Result

Award amounts:

• 1st prize - $2000

• 2nd prize - $1500

• 3rd prize - $1000