Crop Prediction System using Machine Learning

This project, developed by a team of four individuals, aims to predict crop yields based on various features using machine learning. We employ five different algorithms to train the model and predict crop yields.

Team Members

• Tanvi kamanuri
• Yagna valkini
• Sandeep Rajanla
• Karthik

Dataset

The dataset used for this project contains the following features:

• State_Name: Name of the state
• Crop_Type: Type of crop
• Crop: Specific crop name
• N, P, K: Soil nutrient levels (in kg/ha)
• pH: Soil pH level
• Rainfall: Annual rainfall (in mm)
• Temperature: Average temperature (in degrees Celsius)
• Area_in_hectares: Cultivation area in hectares
• Production_in_tons: Crop production in tons
• Yield_ton_per_hec: Yield per hectare (target variable)

Algorithms

We have implemented the following five machine learning algorithms:

1. Random Forest
2. Support Vector Machine (SVM)
3. Decision Tree
4. Gradient Boosting
5. K-Nearest Neighbors (KNN)

Explore the Jupyter notebook Crop_Prediction.ipynb for data analysis and model training.

Results

The results of each algorithm can be found in the Jupyter notebook Crop_Prediction.ipynb file.

Training Results

Algorithm	Desicion Tree Classifier	Random Forest Classifier	KNN	SVM	XGB
train_accuracy	99.998748	99.998748	10.462074	97.717798	99.372801
train_precision	99.998748	99.998748	1.756034	97.853954	99.3849
train_recall	99.998748	99.998748	10.462074	97.717798	99.372801
train_f1	99.998748	99.998748	2.9996	97.756293	99.37602

Testing Results

Algorithm	Desicion Tree Classifier	Random Forest Classifier	KNN	SVM	XGB
test_accuracy	98.452679	98.908363	98.242364	97.651477	98.863295
test_precision	98.45159	98.918539	98.264676	97.834533	98.875949
test_recall	98.452679	98.908363	98.242364	97.651477	98.863295
test_f1	98.451839	98.911897	98.2457	97.698144	98.867799