This project provides extensive information on crop production in India over several years. The ultimate goal is to leverage the data to predict crop production using powerful machine learning techniques.
The content is sourced from the data.world website. You can find the dataset here.
Predicting crop production in India is a crucial task with significant implications for the economy, food security, and policy-making. By developing accurate predictive models, we can help in planning and decision-making processes that affect millions of people.
- Source: https://www.kaggle.com/datasets/abhinand05/crop-production-in-india
- Time Range: 1997-2015
- Geographical Coverage: India
For predicting crop production, the following tools and technologies were used:
- MLflow: For tracking and managing the machine learning experiments.
- DVC (Data Version Control): For versioning the datasets and machine learning models.
- Scikit-learn: For implementing machine learning algorithms.
- Flask: For deploying the machine learning model as a web application.
- Gradio: For creating interactive interfaces to showcase model predictions.
- Download the Dataset: Download the dataset from https://www.kaggle.com/datasets/abhinand05/crop-production-in-india.
- Set Up Environment: Ensure you have the necessary tools installed. You can use the following commands to install them:
pip install mlflow dvc scikit-learn flask gradio
- Explore and Preprocess Data: Load the dataset and perform necessary preprocessing steps.
- Train and Track Models: Use MLflow to track your experiments and train your models using Scikit-learn.
- Version Control with DVC: Use DVC to version control your datasets and models.
- Deploy with Flask: Deploy your trained model using Flask to create a web application.
- Interactive Interface with Gradio: Create an interactive interface for your model predictions using Gradio.
Contributions to improve the dataset and predictive models are welcome. Please feel free to submit pull requests or open issues for discussion.