https://github.com/mehboob14/nextdayrainprediction

As part of a course project in Data Science, I worked on the Nextday Rain Predictions dataset sourced from Kaggle. The project involved end-to-end processes, including data preprocessing, handling class imbalance, and model evaluation. I explored various models such as Logistic Regression and Random Forest, applied appropriate preprocessing techniq
https://github.com/mehboob14/nextdayrainprediction

data-science data-science-projects jupyter-notebooks logistic-regression preprocessing python random-forest

Last synced: 2 months ago
JSON representation

As part of a course project in Data Science, I worked on the Nextday Rain Predictions dataset sourced from Kaggle. The project involved end-to-end processes, including data preprocessing, handling class imbalance, and model evaluation. I explored various models such as Logistic Regression and Random Forest, applied appropriate preprocessing techniq

• Host: GitHub
• URL: https://github.com/mehboob14/nextdayrainprediction
• Owner: mehboob14
• License: mit
• Created: 2025-01-18T11:03:39.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2025-02-06T07:36:31.000Z (over 1 year ago)
• Last Synced: 2025-02-06T08:33:53.695Z (over 1 year ago)
• Topics: data-science, data-science-projects, jupyter-notebooks, logistic-regression, preprocessing, python, random-forest
• Language: Jupyter Notebook
• Homepage:
• Size: 8.18 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
Nextday Rain Predictions

This repository contains the implementation of a Data Science course project focused on rain prediction using the Netday Rain Predictions dataset sourced from Kaggle. The project includes data preprocessing, handling class imbalance, model implementation, and evaluation.

Table of Contents

    Overview

    Dataset

    Technologies Used

    Project Workflow

    Installation and Setup

    Usage

    Results

    Contributing

    License

Overview

The project demonstrates the end-to-end data science pipeline, from data preprocessing to model evaluation. Various techniques were applied to improve prediction accuracy, including:

    Data preprocessing

    Addressing class imbalance

    Feature engineering

    Model training and evaluation

Models explored:

    Logistic Regression

    Random Forest

Dataset

The dataset used for this project is the Netday Rain Predictions dataset, which is available on Kaggle. It includes features that help predict rain events based on historical weather data.

Files in this Repository:

    *.ipynb: Jupyter notebooks containing the step-by-step implementation of the project.

    *.csv: Dataset files used in this project.

Technologies Used

    Python: Core programming language

    Pandas: For data manipulation and analysis

    NumPy: For numerical computations

    Scikit-learn: For implementing machine learning models

    Matplotlib/Seaborn: For data visualization

    Jupyter Notebook: For interactive development

Project Workflow

The workflow of this project includes:

    Data Preprocessing

        Handling missing values

        Encoding categorical features

        Scaling numerical features

    Class Imbalance

        Techniques such as oversampling and undersampling were used.

    Feature Selection

        Correlation analysis and feature importance methods.

    Model Training

        Logistic Regression

        Random Forest

    Model Evaluation

        Metrics such as accuracy, precision, recall, F1-score, and ROC-AUC.

Installation and Setup

    Clone the repository:

git clone https://github.com/your-repo/netday-rain-predictions.git

cd netday-rain-predictions

Create a virtual environment (optional but recommended):

python -m venv env

source env/bin/activate  # On Windows: env\Scripts\activate

Install the required dependencies:

pip install -r requirements.txt

Launch the Jupyter Notebook:

    jupyter notebook

Usage

    Open the Jupyter notebooks to view the step-by-step implementation.

    Place the dataset files (*.csv) in the same directory as the notebooks or update the file paths in the code.

    Run each cell sequentially to reproduce the results.

Results

    The Random Forest model outperformed Logistic Regression in terms of accuracy and precision.

    Key metrics:

        Accuracy: 85%

        Precision: 82%

        Recall: 79%

        ROC-AUC: 88%

Contributing

If you'd like to contribute to this project, feel free to submit a pull request. For major changes, please open an issue first to discuss the changes.

License

This project is licensed under the MIT License. See the LICENSE file for details.