Dashboard Demo de una muestra de sitios webs

EPI 553 HW 3, reviewing Multiple Linear Regression, Tests of Hypotheses, and Interaction Analysis

This project presents the implementation of various clustering methods for customer segmentation based on spending behavior. The analysis uses a dataset containing customer income, purchasing activity, and product expenditure to identify distinct customer groups. Several clustering algorithms, including K-Means, K-Median, DBSCAN, Mean Shift, and Fuzzy C-Means, are applied and compared to evaluate their performance. The results show that customers can be grouped into low, medium, and high-value segments, with each method producing different clustering characteristics. This study highlights the effectiveness of clustering techniques in understanding customer behavior and supporting data-driven marketing strategies.

This document presents the R code and output for a comparative clustering analysis of 34 Indonesian provinces using K-Means, K-Medoids (PAM), DBSCAN, Mean Shift, and Fuzzy C-Means methods based on 16 socioeconomic indicators.

Melakukan studi kasus untuk menganalisis ketidakpastian estimasi rata-rata populasi dengan menggunakan interval kepercayaan 95%.

Proyek ini melakukan segmentasi pelanggan berdasarkan perilaku pengeluaran menggunakan dataset Customer Segmentation Based on Spending Behavior. Beberapa metode clustering diterapkan, yaitu K-Means, K-Median, DBSCAN, Mean Shift, dan Fuzzy C-Means, untuk mengelompokkan pelanggan ke dalam segmen bernilai rendah, menengah, dan tinggi. Hasil analisis menunjukkan bahwa K-Means menghasilkan cluster terbaik berdasarkan evaluasi Silhouette, sedangkan Fuzzy C-Means menawarkan fleksibilitas tambahan untuk data yang overlap. Visualisasi cluster menggunakan PCA dan evaluasi kualitas cluster juga disertakan, sehingga proyek ini dapat menjadi acuan strategi pemasaran berbasis data.