Inspiration: CropAI was inspired by the importance of crop monitoring in agriculture and by the idea of using machine learning to support real-world decision-making.
What it does: CropAI predicts whether a wheat crop is healthy or unhealthy using agricultural input data and displays the result with a confidence score and visual explanations.
How we built it: We built CropAI by preprocessing the dataset, filtering it to wheat samples, training classification models such as Random Forest and Logistic Regression, and integrating the saved model into a Flask website.
Challenges we ran into: Some of the main challenges were improving model performance, dealing with class imbalance, and working without temporal data, which limited the model to one-time crop measurements.
Accomplishments that we're proud of: We are proud of creating a complete machine learning web application that connects data preprocessing, model prediction, and a user-friendly interface in one project.
What we learned: We learned how to clean agricultural data, train and evaluate machine learning models, and make predictions more understandable through a web-based dashboard.
What's next for CropAI: Next, we would like to improve the model, include temporal data, deploy the website online, and expand the system to support more crop types.
Log in or sign up for Devpost to join the conversation.