• 

🌾 EcoPredict – AI-Driven Crop Price Forecasting System 🚩 About the Project

Farmers often sell their crops at low prices due to a lack of access to real-time market information and price prediction tools. This leads to unstable income and economic inefficiencies across the agricultural supply chain.

EcoPredict was created to bridge this gap by providing farmers and traders with an intelligent platform that predicts crop prices and demand trends using machine learning and historical market data. By transforming raw economic data into actionable insights, the platform helps users decide when and where to sell their crops for maximum profit.

💡 Inspiration

The idea for EcoPredict came from observing how small farmers rely on middlemen and often lack awareness of price fluctuations across different markets. Despite agriculture being a major part of the economy, many farmers still make selling decisions based on guesswork rather than data.

We were inspired to build a solution that applies computer science, artificial intelligence, and data analytics to solve a real economic problem affecting millions of people.

🛠️ How We Built It

The system consists of three major components:

1. Data Collection

We gathered historical crop price data from public agricultural market datasets. This data included:

Market location Crop type Historical prices Seasonal trends

1. Machine Learning Model

We implemented a regression-based machine learning model to predict future crop prices.

The prediction model is based on the function:

𝑃 𝑡 +

1

𝑓 ( 𝑃 𝑡 , 𝐷 𝑡 , 𝑆 𝑡 , 𝑊 𝑡 ) P t+1 ​

=f(P t ​

,D t ​

,S t ​

,W t ​

)

Where:

𝑃 𝑡 P t ​

= Current price 𝐷 𝑡 D t ​

= Demand at time t 𝑆 𝑡 S t ​

= Supply at time t 𝑊 𝑡 W t ​

= Weather conditions

This allows the system to estimate future prices based on multiple economic factors.

1. Web Application

We developed a user-friendly web interface where users can:

Select crop type View predicted prices Compare different market locations Visualize trends through charts and graphs 🧰 Technology Stack

Frontend

HTML, CSS, JavaScript

Backend

Node.js / Java Spring Boot

Machine Learning

Python, Scikit-learn

Database

MySQL 📚 What We Learned

Through this project, we gained hands-on experience in:

Integrating machine learning models with web applications Cleaning and processing real-world economic datasets Designing systems that are usable by non-technical users Understanding how economic indicators influence pricing trends

We also learned the importance of data quality and feature selection in building accurate predictive models.

⚠️ Challenges We Faced

1. Data Inconsistency

Agricultural datasets contained missing and inconsistent entries, which required significant preprocessing and cleaning before model training.

1. Model Accuracy

Early versions of the model produced unstable predictions due to seasonal price volatility. We improved accuracy by incorporating additional features and normalizing the data.

1. System Integration

Connecting the machine learning model with the backend and ensuring real-time predictions in the web interface required careful API design and testing.

🚀 Impact and Future Scope

EcoPredict has the potential to:

Increase farmer income Reduce dependency on middlemen Improve price transparency in agricultural markets

In the future, we plan to:

Integrate real-time market APIs Add multilingual support for rural accessibility Deploy the system as a mobile application for wider adoption 🏁 Conclusion

EcoPredict demonstrates how computer science and economics can work together to create practical, data-driven solutions. By empowering farmers with predictive insights, we aim to contribute to a more efficient and equitable agricultural economy.

Built With

• html5
• java
• python
• spring