The project aims to utilize a particular crop related dataset (Paddy / Rice Dataset from UC Irvene Machine Learning repository utilized in current case), which contains multiple agronomic, environmental, and crop‑related features, for the purpose of dimension reduction. Modern agricultural research increasingly relies on large, feature‑rich datasets to understand crop performance, optimize cultivation practices, and support data‑driven decision‑making. As farming conditions, climate patterns, and crop varieties evolve, the volume and complexity of agricultural yield continues to grow which is an expected practice. For current project, we utilize the full Paddy Dataset because all feature groups—soil characteristics, climate variables, crop breed or traits, and management practices—contribute to understanding paddy or rice growth patterns. Small variations across a few selective features can significantly influence yield, making dimension reduction a valuable tool for uncovering underlying structure in the dataset. And, accordingly the results can be utilized to harness parameters which influence paddy production volume the most for real world cultivation suggestions.

Clustering Night Light Intensity related data extracted using NOAA/DMSP-OLS/NIGHTTIME_LIGHTS dataset available Google Earth Engine end. The data can be used for identifying urban centers for future planning and disaster response purpose. Also, newer datasets from Google Earth Engine (available under commercial version) can be fetched and added to the paper as a part of future scope.