This project focuses on predicting river water discharge for River Kabul at the Nowshera Bridge using key hydrological parameters, specifically rainfall and temperature. It evaluates and compares various machine learning algorithms and identifies the most suitable model for accurate discharge forecasting.
This work has been carried out as part of a BS (Computer Science) research project by the following students at the University of Engineering & Technology (UET), Peshawar:
- Naqeeb Ullah
- Omais Ahmad Khan
- Muhammad Zohaib
The project was guided and supervised by:
Dr. Laeeq Ahmed
Assistant Professor, Department of Computer Science
UET Peshawar, Pakistan
PhD in Computational Science and Technology, KTH, Sweden
laeeq80@gmail.com
LinkedIn
Google Scholar
_Note: Dr. Laeeq Ahmed served as the Research Lead for this project.
The objective of this project is to develop a predictive model for forecasting river discharge based on meteorological inputs. Key features include:
- Input Parameters: Rainfall and Temperature
- Output: Predicted Discharge (m³/s) at the Nowshera Bridge
- Algorithms Analyzed: Multiple machine learning models were implemented and compared.
- Findings: Among all tested algorithms, the Artificial Neural Network (ANN) outperformed others in terms of prediction accuracy.
- Python 3.x
- pandas, numpy
- scikit-learn
- matplotlib, seaborn, joblib
- Keras (for ANN modeling)
- Jupyter Notebook / vsCode
git clone https://github.com/laeeq80/riverFlowPrediction.git cd riverFlowPrediction Open project in an IDE e.g. vsCode and run the current file. ## Results Summary Algorithms Evaluated: Linear Regression, Decision Trees, Random Forest, Support Vector Regression, Artificial Neural Networks Performance Metrics: RMSE, R², MAE Best Model: ANN demonstrated the highest accuracy and reliability for river discharge prediction ## License This repository is provided for academic and educational purposes only. Reuse or adaptation for other projects should properly acknowledge the student authors and supervising faculty. ## Contact For questions or collaboration inquiries: Dr. Laeeq Ahmed – laeeq80@gmail.com