Inspiration
The agricultural industry is one of the most essential industries in the world, yet it remains one of the least advanced when it comes to using data and intelligence for efficiency. With the growing population and ever-decreasing number of people getting into agriculture and farming it has become vital to increase the output of farming sustainably. Growing the right crop at the right place at the right time will be the most important factor in the coming years for sustainable development
What it does
The proposed system will predict the most suitable crop for particular land based on soil contents and weather parameters such as Temperature, Humidity, soil PH It defines the target for a model. After data cleaning the dataset will be split into training and test set. This system helps farmers in agriculture. It helps in soil classification and suggest crop based on type of soil and weather conditions.
How we built it
The Random Forest model for the crop recommendation system was built by first preparing the dataset, which included cleaning data, encoding categorical variables, and splitting it into training and testing sets. The model was trained using Random Forest, an ensemble learning algorithm that constructs multiple decision trees to predict the most suitable crop based on agricultural features like soil type, temperature, and humidity. After training, hyperparameters were fine-tuned to optimize performance, and the model was evaluated using accuracy and other metrics. Finally, the trained model was deployed for making predictions on new input data for crop recommendations.
Challenges we ran into
We developed a crop recommendation system by analyzing an agricultural database, focusing on soil properties, nutrients, and environmental factors while addressing data inconsistencies. Key features like Nutrient Index and Climate Compatibility were engineered to enhance model performance. Machine learning models, including Random Forest and Gradient Boosting, were trained, tackling imbalanced data with SMOTE. Metrics such as accuracy and F1-score ensured reliable multi-class predictions. A user-friendly dashboard was created for farmers to input conditions and receive crop recommendations. Clear insights explained predictions, linking crop suitability to soil, nutrients, and climate factors.
What we learned
we learnt working on different tools and real world datasets which equips us with hands on experience and helps us to gain knowledge and work on different industrial problems and goals
Log in or sign up for Devpost to join the conversation.