Overview / Usage

This project aims to predict freshwater quality using an adaptive particle swarm optimization algorithm with the convolutional neural network. The goal is to provide accurate predictions for monitoring water quality and identifying potential risks to human health. The adaptive PSO algorithm is used for feature selection to improve the accuracy of the CNN model. While the results show promising improvements, more work is needed for production-level implementation. This research can be used by water quality management authorities and environmental organizations to improve their monitoring and management practices.

Methodology / Approach

Thee methodology used involved utilizing adaptive particle swarm optimization (APSO) algorithm to select the most relevant features from the dataset, followed by the development of a convolutional neural network (CNN) model to predict freshwater quality. The framework for developing the CNN model was TensorFlow, a widely used open-source software library for machine learning and artificial intelligence. The model was trained using the backpropagation algorithm, with optimization performed using the Adam optimizer. Dropout regularization was applied to prevent overfitting, and the model was evaluated using metrics such as accuracy and loss. Overall, the methodology involved using a combination of feature selection and deep learning techniques to solve the problem of predicting freshwater quality.

Technologies Used

This work's development involves using several technologies, libraries, tools, and software. These include:

1. Intel technologies: This project uses Intel's oneAPI Python library, a unified programming model for various architectures, including CPUs, and GPUs.
2. Scikit-learn: A machine learning library in Python, which is used for preprocessing, feature selection, and training XGBoost models.
3. TensorFlow: An open-source machine learning library in Python, which is used for training and evaluating Convolutional Neural Network (CNN) models.
4. Anaconda: A popular open-source distribution of Python and R programming languages, which is used to manage dependencies and create a virtual environment for the project.
5. Jupyter Notebook: An open-source web application that allows the creation and sharing of documents that contain live code, equations, visualizations, and narrative text.
6. NumPy and Pandas: Python libraries used for data manipulation and analysis.
7. XGBoost: A scalable and efficient implementation of gradient boosting for decision trees, used for classification in this project.
8. Adaptive Particle Swarm Optimization (APSO): A metaheuristic optimization algorithm used to perform feature selection in the XGBoost model.

Repository

https://github.com/Nitin-Mane/Freshwater-Quality-Prediction-using-oneAPI