This project, developed during the Statistical Inference course at FCIS, Ain Shams University, employs R to conduct customer segmentation analysis. Various clustering algorithms, including k-means, DBSCAN, and hierarchical clustering, were implemented to derive meaningful insights from the dataset. The project focuses on data preprocessing, exploratory data analysis, and visualization techniques to uncover patterns in customer behavior.

• Dataset
• Data Preprocessing
• Exploratory Data Analysis
• Clustering Algorithms
  • K-means
  • DBSCAN
  • K-medoids
  • Hierarchical Clustering
• Results
• Contributing

The dataset, sourced from "CustomerSegmentation.csv," contains information about customers, including age, annual income, and spending score.

The dataset underwent thorough preprocessing steps, including handling missing values, removing negative values, and encoding categorical data. Descriptive statistics and visualizations were utilized to ensure data quality.

Exploratory Data Analysis (EDA) involved visualizing gender distribution, age distribution, and analyzing annual income and spending score patterns. These insights formed the basis for subsequent clustering.

The optimal number of clusters was determined using the elbow method, silhouette method, and gap statistic. The K-means algorithm was then applied to segment customers.

DBSCAN clustering was performed with an estimated epsilon value, visualizing the clusters and their characteristics.

PAM algorithm was employed to estimate the optimal number of clusters, and K-medoids clustering was visualized.

Hierarchical clustering using both hclust and agnes methods was executed, and a dendrogram was plotted.

The project resulted in distinct customer segments identified through various clustering algorithms, providing valuable insights for business decision-making.

Contributions are welcome! Fork the repository and create a pull request with your enhancements.