https://github.com/jatin-mehra119/roe-prediction-modeling

A Web-APP for predicting Return of Equity using Machine learning model.
https://github.com/jatin-mehra119/roe-prediction-modeling

data-analysis data-science forecasting machine-learning regression-models scikit-learn

Last synced: about 19 hours ago
JSON representation

A Web-APP for predicting Return of Equity using Machine learning model.

• Host: GitHub
• URL: https://github.com/jatin-mehra119/roe-prediction-modeling
• Owner: Jatin-Mehra119
• License: cc0-1.0
• Created: 2024-07-24T11:16:16.000Z (4 months ago)
• Default Branch: main
• Last Pushed: 2024-07-24T13:29:29.000Z (4 months ago)
• Last Synced: 2024-07-25T13:38:30.189Z (4 months ago)
• Topics: data-analysis, data-science, forecasting, machine-learning, regression-models, scikit-learn
• Language: Jupyter Notebook
• Homepage: https://roe-prediction-jatinmehra119.streamlit.app/
• Size: 10.7 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: readme.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Return On Equity Prediction Project

This project aims to predict the Return on Equity (ROE) of companies based on various financial metrics using a machine learning model. The project includes data analysis, model training, and a web application for making predictions.

#### Data source - [Link](https://dataverse.harvard.edu/dataset.xhtml;jsessionid=41510459a96a8d0e58860d844857?persistentId=doi:10.7910/DVN/XMVD5L)

#### App - [link](https://roe-prediction-jatinmehra119.streamlit.app/)

 - Notebook.ipynb - It includes all the EDA, Preprocessing, Model Training and Model Evaluation. 

 - ROE_model.pkl - Pipelined Model Saved as a pickle file.  

 - app.py - Streamlit app used for making prediction using pickled model.

### Key Insights gained more training the model-

1.  **Top Contributors**:

    

    -   **Equity** (`log__Equity`): The most important feature, suggesting that the equity of a company is a significant predictor of its ROE.

    -   **Current Ratio** (`log__Current_Ratio`): Indicates the liquidity of a company and its ability to pay short-term obligations, which is a strong predictor.

    -   **Market Cap** (`log__Market_Cap`): Reflects the company's size and market perception, contributing substantially to predicting ROE.

    -   **Profit/Loss** (`log__Profit_Loss`): Indicates the profitability of the company, also a significant predictor.

2.  **Moderate Contributors**:

    

    -   **Assets** (`log__Assets`): Although lower than the top contributors, still has a meaningful impact.

    -   **Dividends** (`log__Dividends`): Reflects the company's dividend policy, showing some importance.

    -   **Liabilities** (`log__Liabilities`): Indicates the company's obligations, but with lesser importance compared to assets and equity.

    -   **Expenses** (`norm__Expenses`): Operating expenses also contribute, but to a lesser degree.

    -   **Growth Rate** (`norm__Growth_Rate`): The company's growth rate is considered, but with minimal importance.

3.  **Low Contributors**:

    

    -   **Revenue by Region** (`log__Revenue_Central China`, `log__Revenue_West China`, `log__Revenue_South China`, `log__Revenue_North China`, `log__Revenue_East China`): These features have very low importance, indicating regional revenues might not significantly influence ROE.

    -   **PE Ratio** (`log__PE_Ratio`): Also shows low importance in predicting ROE.

    -   **Total Revenue** (`norm__Revenue`): Surprisingly low impact on predicting ROE.

4.  **Minimal or Negligible Contributors**:

    

    -   **City and Sector**: Categorical variables like `City` and `Sector` have very low importance.

    -   **Company Size**: The size of the company (`onehot__Company_Size_Medium`, `onehot__Company_Size_Small`, `onehot__Company_Size_Large`) has almost no importance in predicting ROE.