Crop recommendation

Published: 24 July 2025| Version 1 | DOI: 10.17632/vynxnppr7j.1
Contributor:
Sagana Thangatamilan

Description

The dataset is collected by integrating multiple data sources such as soil properties (type of soil, soil pH), climate factors (temperature, relative humidity, season), nutrient levels (N (nitrogen), P (phosphorus), K (potassium)), crop characteristics (crop duration, water required, type of crop), and agricultural factors (sown, harvested, water source).

Files

Steps to reproduce

The dataset is collected by integrating multiple data sources such as soil properties (type of soil, soil pH), climate factors (temperature, relative humidity, season), nutrient levels (N (nitrogen), P (phosphorus), K (potassium)), crop characteristics (crop duration, water required, type of crop), and agricultural factors (sown, harvested, water source). Data is collected from multiple sources in different agricultural websites of Tamil Nadu. From the raw data, a new dataset is generated using CTGAN, preprocessing is done with One-hot encoding and mode-specific normalization, feature selection techniques are used to select the most important features that mimic the crop recommendation. After features are extracted, the data is fed into machine learning models, and the performance of each model is measured using Accuracy, Precision, Recall, and F1-Score, and a final recommendation for the specific land is made based on the input provided to the model.

Institutions

  • Kongu Engineering College

Categories

Classification System, Recommendation System, Crop Characteristics

Licence