Rainfall Prediction System
Overview
The Rainfall Prediction System is a machine learning-based project aimed at providing accurate rainfall predictions for the regions of Vavuniya, Anuradhapura, and Maha Illuppallama. The system utilizes ensemble methods and individual machine learning models to forecast rainfall patterns, assisting farmers and resource managers in making informed decisions related to agriculture and water management.
Benefits
- Improved Agricultural Productivity: Optimizes irrigation schedules and crop selection, leading to increased crop yields.
- Water Resource Management: Enables efficient planning of water resources, reservoirs, and irrigation systems.
- Risk Mitigation: Proactively mitigates risks such as crop failure and water scarcity through reliable predictions.
- Cost Savings: Reduces resource wastage, resulting in cost savings for farmers and agricultural businesses.
- Environmental Sustainability: Promotes sustainable agricultural practices, minimizing environmental impact.
- Empowerment of Stakeholders: Provides actionable information to farmers, extension workers, and government agencies.
- Resilience to Climate Change: Enhances resilience to changing environmental conditions and weather variability.
- Community Engagement: Fosters collaboration, knowledge sharing, and capacity building within the community.
- Socioeconomic Impact: Contributes to long-term socioeconomic development and poverty alleviation in rural communities.
Features
- Data Collection: Historical rainfall data is collected from the Meteorological Department in Sri Lanka to build predictive models.
- Data Preprocessing: Exploratory Data Analysis (EDA) techniques are applied to preprocess the collected data, ensuring its quality and relevance.
- Model Selection: Various machine learning algorithms are employed, including SARIMA, decision tree, random forest, Gradient Boosting Machine (GBM), XGBoost, and Long Short-Term Memory (LSTM) networks.
- Evaluation Metrics: Each model is evaluated using metrics such as Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and R-value to assess its predictive performance.
Installation
- Clone the repository:
https://github.com/cepdnaclk/e19-co544-Rainfall-Prediction-System
Evaluation
- Project Proposal : https://www.canva.com/design/DAGDVjEfCWo/Tof1LMkaLnMqujyno09xag/edit
- Mid Evaluation : https://www.canva.com/design/DAGGiM-CoMo/EhQWItXYlVBEMJ8a7_FIVA/edit
Team
- E/19/060, Danujan, email
- E/19/131, Kasuni, email
- E/19/205, Kumara I.P.S.N.U., email
- E/19/249, Muthukumarana M.P.S.A. email
- E/19/266, Nithusikan, email
Links