research · 2024
Rice Production Clustering
Research and analysis
- RapidMiner
- K-Means
Summary
A clustering study over 27 years of rice production data for Sumatra, from 1993 to 2020.
Approach
K-Means clustering in RapidMiner, with the cluster count chosen to keep the groups interpretable for a policy reader.
Outcome
Three productivity clusters that separate the region's production behaviour over time.
Honest limitations
Clusters describe the historical record. They are not a forecast.
Gallery
Next project
TOPSIS vs SMART