Water Level Distribution

Project for the Deep Learning course year 2024/2025

Problem Statement

Accurate prediction of water levels is crucial for preventing floods and managing water resources. Traditional methods may not provide the required accuracy, which is where this project steps in by leveraging data-driven approaches.

Data

The dataset consists of historical water level readings and associated features that influence these levels. Data cleaning and preprocessing are pivotal for improving model performance.

Approach

The project employs a variety of machine learning models, including regression techniques and neural networks, to achieve accurate predictions. The approach also involves feature selection and hyperparameter tuning to enhance model performance.

How to Run (Google Colab)

To run this project in Google Colab:

1. Open the provided Jupyter Notebook in Google Colab.
2. Ensure that you have access to the necessary dataset.
3. Follow the instructions in the Notebook to execute the code cells and visualize the results.

Repository Structure

• Water_level_prediction_last__Esther_Giuliano.ipynb: Jupyter Notebook containing the implementation
• data/: Directory containing the datasets used in the project
• models/: Directory for storing model scripts
• results/: Directory for storing results and visualizations

Results

The results indicate significant improvements in the predictive capabilities for water levels compared to baseline methods, highlighting the effectiveness of machine learning in this domain.