kasunUdayanga/Tea_Yield_Prediction_Model

Tea Yield Prediction Model 🌱

Model Description

This is a Linear Regression model that predicts tea crop yield (in kg/ha) using six key agricultural and environmental factors. The model was selected as the best performer among four algorithms tested, achieving an R² score of 0.6448.

Key Features

• ✅ Best performer among Linear Regression, Decision Tree, Random Forest, and SVR
• ✅ Simple & interpretable linear model
• ✅ Practical application for agricultural planning
• ✅ Ready-to-use with minimal dependencies

Model Performance

Metric	Value	Description
R² Score	0.6448	Explains 64.48% of yield variance
MAE	200.27 kg/ha	Average prediction error
RMSE	254.21 kg/ha	Error with penalty for large mistakes
Training Samples	47,536	After preprocessing
Features	6	Agricultural/environmental factors

Input Features

Feature	Type	Range	Description
rainfall_mm	float	50-220 mm	Monthly rainfall
temperature_avg	float	18-30°C	Average temperature
soil_ph	float	4.5-6.0	Soil pH level
fertilizer_kg_ha	float	200-500	Fertilizer application rate
plant_age_years	float	2-25	Age of tea plants
altitude_m	float	500-2000	Farm elevation

Quick Start

Installation

pip install scikit-learn pandas joblib