Abstract
Agricultural productivity is significantly influenced by various environmental factors such as temperature, rainfall, humidity, wind speed, and soil conditions. Traditional farming practices often rely on historical data and farmer experience, which may not be sufficient to meet the growing demand for food in a changing climate. This project aims to develop a real-time crop yield prediction system using machine learning techniques. By integrating live weather data through APIs and historical crop performance datasets, the system predicts the most suitable crop to cultivate based on current environmental conditions. The model employs supervised learning algorithms—such as Decision Trees or Random Forests—to analyze patterns and recommend crops with high yield potential. The real-time prediction feature not only enhances decision-making for farmers but also contributes to sustainable agriculture by optimizing resource use. A user-friendly graphical interface ensures ease of access and usability, making the system a practical tool for precision farming and agritech solutions.
Keywords : dataset of weather, machine learning









Introduction
Agriculture plays a vital role in the economy and food security of nations around the world. The growing population and changing climatic conditions have created a demand for more efficient and sustainable farming methods. Crop yield is influenced by a variety of environmental and soil-based factors, including temperature, humidity, rainfall, wind speed, and soil moisture. However, farmers often rely on traditional knowledge or past experiences, which may not always provide the most accurate guidance in today’s rapidly changing climate. The emergence of machine learning and artificial intelligence has opened new opportunities in the agricultural sector. These technologies allow for the development of predictive systems that can analyze large datasets and provide accurate forecasts. In particular, crop yield prediction using machine learning has gained traction due to its ability to learn patterns from historical data and apply them to real-time inputs. This approach not only improves decision-making for farmers but also minimizes the risks associated with unpredictable weather and market fluctuations.
This project focuses on the implementation of a real-time crop yield prediction system using machine learning algorithms. The model is trained on a dataset that includes various weather parameters and corresponding crop yields. Once trained, the system can take live weather data from APIs such as Weatherstack and predict the most suitable crop for the given environmental conditions. This ensures that the recommendations are current, accurate, and region-specific, making them highly practical for real-world application. The key features of this system include its ability to work in real-time, its use of supervised learning models (such as Decision Trees or Random Forests), and its integration with live data sources. By continuously updating the input data, the system maintains its relevance and accuracy, which is crucial for agricultural planning. Additionally, the system supports automated encoding of categorical data such as district names and wind directions, ensuring seamless compatibility with the machine learning model.
To enhance usability, the system includes a graphical user interface (GUI) that allows users to easily input or view data without needing to interact directly with the code. This GUI displays weather information, model inputs, and the recommended crop in a clear and user-friendly format. Farmers or agricultural officers can use this interface to make quick decisions based on current weather conditions, even without a background in data science.In conclusion, the proposed real-time crop yield prediction system represents a practical and scalable solution to modern agricultural challenges. By leveraging machine learning and live data, it offers dynamic, intelligent crop recommendations that can help increase productivity, reduce uncertainty, and contribute to more sustainable farming practices. This approach bridges the gap between traditional agriculture and modern technology, empowering farmers with tools to make informed and timely decisions.


Objective
The main objective of this project is to develop a real-time crop yield prediction system using machine learning algorithms that can analyze live environmental data to recommend the most suitable crops. The system is designed to enhance agricultural productivity, reduce crop failure, and assist farmers in informed decision-making. The specific objectives include:
1. To build a machine learning model that can accurately predict crop yield or recommend the best-suited crop based on real-time weather and environmental parameters.
2. To utilize live weather data through integration with APIs (e.g., Weatherstack) for dynamic and up-to-date decision-making.
3. To preprocess and analyze agricultural datasets, including features like temperature, humidity, rainfall, wind speed, wind direction, and soil characteristics.
4. To encode categorical variables such as district names and wind directions using label encoding or similar techniques for compatibility with machine learning models.
5. To design a user-friendly graphical user interface (GUI) that allows users (e.g., farmers, agricultural officers) to input location data and receive crop recommendations in real-time.
6. To test and validate the system’s performance using real-world data to ensure accuracy, scalability, and usability in diverse agricultural conditions.