https://github.com/sevilaymuni/lgbm-homicide-prediction-app

LightGBM ML application deployment via Streamlit.
https://github.com/sevilaymuni/lgbm-homicide-prediction-app

crime-analysis data-science deployment gradient-boosting lightgbm machine-learning predictive-modeling streamlit streamlit-webapp

Last synced: 7 days ago
JSON representation

LightGBM ML application deployment via Streamlit.

• Host: GitHub
• URL: https://github.com/sevilaymuni/lgbm-homicide-prediction-app
• Owner: SevilayMuni
• Created: 2024-11-04T13:06:38.000Z (3 months ago)
• Default Branch: master
• Last Pushed: 2024-11-24T13:10:04.000Z (3 months ago)
• Last Synced: 2024-12-16T03:19:44.826Z (2 months ago)
• Topics: crime-analysis, data-science, deployment, gradient-boosting, lightgbm, machine-learning, predictive-modeling, streamlit, streamlit-webapp
• Language: Jupyter Notebook
• Homepage: https://sevilaygirgin-app-homicide-prediction.streamlit.app/
• Size: 10.4 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# 🕵🏻 Justice Forecast: Solvability Analysis for Homicide Cases 👩🏻‍💻

## The project is dedicated to all murder victims and their families whose justice has not been served yet.

The project aims to conduct data science research and demonstrate the importance of accurately accounting for unsolved homicides within communities. 

The model data source is [Murder Accountability Project](https://www.murderdata.org/)

## Check Out My Application

[![My Application](https://img.shields.io/badge/Streamlit-FF4B4B?style=for-the-badge&logo=Streamlit&logoColor=white)](https://sevilaygirgin-app-homicide-prediction.streamlit.app/)

[![My Medium Article](https://img.shields.io/badge/Medium-12100E?style=for-the-badge&logo=medium&logoColor=white)](https://medium.com/@sevilaymuniregirgin/justice-forecast-app-for-homicide-solvability-43536fef50ad)

## Model Evaluation

*Feature importance plot based on **gain**. It quantifies the model’s accuracy improvement achieved by using specific features for splitting.*

[](https://github.com/SevilayMuni/LGBM-homicide-prediction-app/blob/master/images2/Gain-Feature-Importance-Plot.png)

*Feature importance plot based on **split**. This measures the number of times a feature is used to split the data across all trees in the model.*

[](https://github.com/SevilayMuni/LGBM-homicide-prediction-app/blob/master/images2/Split-Feature-Importance-Plot.png)

All in all, **important features for homicide solvability**: 

    Relationship between the victim and offender

    The circumstance (or theory) of the crime

    Year of the homicide

    Victim's Age

## Variable Description Table

Feature | Description |

-----|-----|

Agentype| Type of the law enforcement agency

Year| Year of homicide (or victim’s body was recovered) 

Month| The month of homicide (or victim’s body was recovered)

Murder| 1: Murder & 0: Negligent Manslaughter

VicAge| Victim’s age

VicSex| Victim’s sex (“Unknown” gender: incomplete remains were recovered)

VicRace| Victim’s race

Weapon| Weapon used in the crime

Relationship| The relationship between victim and offender

Circumstance| The circumstances (or theory) of the crime

VicCount| Victim number in the crime

Region| USA region in which the homicide was reported

## Contact 📩

For any questions or inquiries, feel free to reach out:

- [**My Website**](https://sevilaymuni.github.io/Girgin/)

- **LinkedIn:** [Sevilay Munire Girgin](www.linkedin.com/in/sevilay-munire-girgin-8902a7159)

Thank you for visiting my project repository. Happy and accurate classification! 💕

| One's destination is never a place but rather a new way of seeing things. - Henry Miller |

-----|

 Copyright (c) 2024 Sevilay Münire Girgin