Summary of Feasibility Of Machine Learning-based Rice Yield Prediction in India at the District Level Using Climate Reanalysis Data, by Djavan De Clercq et al.
Feasibility of machine learning-based rice yield prediction in India at the district level using climate reanalysis data
by Djavan De Clercq, Adam Mahdi
First submitted to arxiv on: 12 Mar 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary Machine learning models were trained on 20 years of data from Indian rice-producing districts to predict Kharif season rice yields at the district level. Nineteen models, including CatBoost, LightGBM, Orthogonal Matching Pursuit, and Extremely Randomized Trees, were used to build a machine learning pipeline that demonstrated reasonable accuracy in predicting rice yields with out-of-sample R2, MAE, and MAPE performance of up to 0.82, 0.29, and 0.16 respectively. These results outperformed related literature on rice yield modeling in other contexts and countries. SHAP value analysis was conducted to infer the importance and directional impact of climate and remote sensing variables included in the model. Important features driving rice yields included temperature, soil water volume, and leaf area index. In particular, higher temperatures in August correlate with increased rice yields, particularly when the leaf area index in August is also high. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Machine learning helps predict how much rice India will harvest before it’s grown! This study used special models to forecast rice yields at the district level several months ahead of time. They used 20 years of data from Indian districts and compared different models like CatBoost, LightGBM, and more. The results showed that these models were really good at predicting rice yields, even better than other studies in similar areas. They also found out which factors, like temperature and soil moisture, are most important for rice growth. This can help farmers and others make better decisions about how much rice to grow and where. |
Keywords
* Artificial intelligence * Machine learning * Mae * Temperature
