Logo Lanfrica

esmael-uta/Car-Insurance-Data-Analysis

Domaine:

socioeconomic

Type de record:

project
Créateur:
esm
Hôte:
This project focuses on analyzing historical car insurance claim data for AlphaCare Insurance Solutions (ACIS) in South Africa. The primary goal is to optimize marketing strategies and identify "low-risk" clients for targeted premium reductions to attract new customers. # AlphaCares Insurance Analytics ### This project focuses on analyzing historical car insurance claim data for AlphaCare Insurance Solutions (ACIS) in South Africa. The primary goal is to optimize marketing strategies and identify "low-risk" clients for targeted premium reductions to attract new customers. This repository is organized into branches, each corresponding to a specific task: - **`task-1`**: Exploratory Data Analysis (EDA) - **`task-2`**: Data Version Control (DVC) - **`task-3`**: A/B Hypothesis Testing - **`task-4`**: Statistical Modeling The `main` branch contains the final, consolidated version of all tasks. --- ## Task 1: Exploratory Data Analysis (EDA) ### Objective To understand the data through summarization, statistical analysis, and visualization. ### Steps Completed 1. **Git Repository Setup**: - Created this repository and initialized Git version control. - Created a branch named `task-1` for EDA. - Made descriptive commits at least three times per day. 2. **Data Summarization**: - Descriptive statistics calculated for numerical features (e.g., `TotalPremium`, `TotalClaims`). - Verified data types and reformatted columns as needed. 3. **Data Quality Assessment**: - Checked for missing values and handled them appropriately. 4. **Univariate and Bivariate Analysis**: - Plotted histograms, bar charts, scatter plots, and correlation matrices. 5. **Outlier Detection**: - Used box plots to detect and cap outliers. 6. **Visualization**: - Produced creative and insightful plots to present key findings. --- ## Task 2: Data Version Control (DVC) ### Objective To implement data version control to track and manage datasets using DVC. ### Steps Completed 1. **Installed DVC**: - Initialized DVC in the project directory. 2. **Configured Local Remote Storage**: - Created a local storage directory and added it as a DVC remote. 3. **Tracked Data**: - Used `dvc add` to track dataset versions. 4. **Version Control Integration**: - Committed `.dvc` …