https://github.com/szymon-budziak/flood_prediction

Flood prediction with Linear Regression and other ML models
https://github.com/szymon-budziak/flood_prediction

Last synced: about 2 months ago
JSON representation

Flood prediction with Linear Regression and other ML models

• Host: GitHub
• URL: https://github.com/szymon-budziak/flood_prediction
• Owner: Szymon-Budziak
• License: mit
• Created: 2024-06-10T12:39:33.000Z (11 months ago)
• Default Branch: main
• Last Pushed: 2024-07-30T16:19:51.000Z (9 months ago)
• Last Synced: 2025-02-01T02:49:48.220Z (3 months ago)
• Language: HTML
• Size: 18.4 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Flood prediction

This project aims to predict flood probability based on various features. The dataset used is from

the [Regression with a Flood Prediction Dataset](https://www.kaggle.com/competitions/playground-series-s4e5) from Kaggle. 

The project was done in Python and R.

## Installation

For the Python part, you need to run the following command:

```bash

poetry install

```

## Data

The comprehensive data description can be found in Python notebook. Some of the features are:

 - MonsoonIntensity: Index representing monsoon intensity. Higher values indicate greater monsoon rainfall intensity. (strong temporary wind)

 - RiverManagement: Index representing the effectiveness of river management. Higher values indicate more effective management.

 - IneffectiveDisasterPreparedness: Index representing the ineffectiveness of disaster preparedness. Higher values indicate less effective preparedness.

 - PopulationScore: Index representing population density. Higher values indicate greater population density.

Target variable

 - FloodProbability: Flood probability, expressed as a continuous value between 0 and 1, where higher values indicate greater flood probability.

## Models comparison

The following plots were created to compare the performance of different models:

 - Numerical features

![Numerical features](src/images/histograms_numerical_features.png)

- MSE of all models

![MSE of all models](src/images/mse_all_models.png)

- R^2 of all models

![R^2 of all models](src/images/r2_all_models.png)