Intelligent Prediction System

Smart Crop Advisor

Enter your soil nutrients and climate conditions. Our KNN-based model will recommend the best crop for your field.

🌱
22 crops
CROP TYPES
📦
2,200 rows
DATASET SIZE
🔬
7 features
INPUT PARAMS
🎯
KNN + RF + DT
ALGORITHMS USED
🧪 Soil & Climate Parameters
N — Nitrogen
mg/kg · 0–140
70
P — Phosphorus
mg/kg · 5–145
50
K — Potassium
mg/kg · 5–205
40
Temperature
°C · 10–45
25
Humidity
% · 14–100
60
pH Value
3.5–9.5
6.5
Rainfall
mm/year · 20–300
150
🌿 Prediction Result
🌾

Adjust parameters and click
Analyze & Predict to get recommendations

🌱
—
—
Match Confidence—
Also Suitable
Data Analysis

Model Analytics

Performance metrics, algorithm comparisons, and dataset insights from the Kaggle Crop Recommendation Dataset.

🌳
8% acc
DECISION TREE
🌲
4% acc
RANDOM FOREST
🔍
13% acc
KNN (BEST)
⚠️
Low note
RANDOM LABELS
Algorithm Accuracy Comparison
METRICS
Crop Distribution in Dataset
100 EACH
Average Nitrogen per Crop
mg/kg
Average Rainfall per Crop
mm/year
Temperature Range by Crop
°C AVG
Dataset Overview

Crop Database

All 22 crop profiles with average nutrient and climate requirements from the 2,200-row Kaggle dataset.

CropType N (avg)P (avg)K (avg) Temp °CHumidity %pHRainfall mm
Nutrient & Climate Profile

Crop Radar View

Compare the nutrient and climate profile of any two crops side by side.

Crop A Profile
Crop B Profile
Project Documentation

About This Project

📂
Data Collection
🧹
Preprocessing
🏗️
Data Warehouse
🤖
ML Training
📊
Evaluation
🌐
Deployment

🎯 Problem Statement

Indian farmers often struggle to choose the right crop due to variations in soil quality and climate. This leads to low yield and financial losses. We built a data-driven solution that uses machine learning to recommend the most suitable crop based on soil nutrients (N, P, K), temperature, humidity, pH, and rainfall.

📦 Dataset

Source: Kaggle — Crop Recommendation Dataset by Atharva Ingle

Records: 2,200 rows · 22 crop types · 7 features

Features: N, P, K (soil nutrients), Temperature, Humidity, pH, Rainfall

🤖 Algorithms Used

🌳

Decision Tree

max_depth=10, criterion='gini'

8% accuracy
🌲

Random Forest

n_estimators=100, random_state=42

4% accuracy
🔍

KNN Classifier

n_neighbors=5, metric='minkowski'

13% accuracy ⭐ Best

🏗️ Data Warehouse Design

Star Schema: Crop_Yield_Fact connected to Soil_Dimension, Climate_Dimension, Crop_Dimension — simple structure with faster query performance.

Snowflake Schema: Normalized extension of Star Schema with sub-dimension tables (Soil_Type, Season, Crop_Type) — reduces data redundancy.

🚀 Future Scope

Real-time soil sensor data integration · Weather API for live climate feed · Mobile app for farmers · Yield quantity prediction using regression models