Customer Clustering Output: Combined Static (RFM) and Temporal (DTW) Segmentation Data

This dataset contains the final multi-dimensional customer segmentation labels generated from a hybrid clustering framework. It combines static value-based segmentation and dynamic behavioral trends to provide a comprehensive profile of customer transactional activity. The dataset includes cluster assignments for 2,194 unique customers, derived from a pipeline that utilizes K-Means clustering for static spatial RFM (Recency, Frequency, Monetary) data and TimeSeriesKMeans with Dynamic Time Warping (DTW) for sequential temporal data. Data Schema & Dictionary The file `clustering_result.csv` consists of the following three columns: `CustomerID`: Unique identifier for each customer (anonymous/masked strings). `Static_Kmeans_Cluster`: Categorical label representing the static RFM segment (0, 1, 2, 3). This ranks customers based on their overall transaction value, frequency, and recency up to the cutoff date. `TS_Cluster`: Categorical label representing the dynamic time-series behavioral segment (0, 1, 2, 3). This groups customers based on the similarity of their weekly purchase trajectories over time using DTW. Dataset Insights & Distribution The distribution of the 2,194 customers across the respective clusters is structured as follows: Static K-Means Clusters: Cluster 0: 892 customers Cluster 1: 1,269 customers Cluster 2: 1 customer Cluster 3: 32 customers Dynamic Time-Series (TS) Clusters: Cluster 0: 2,135 customers Cluster 1: 2 customers Cluster 2: 1 customer Cluster 3: 56 customers Context & Methodology This dataset serves as the direct output of the `clustering.py` processing pipeline. By cross-tabulating `Static_Kmeans_Cluster` and `TS_Cluster`, researchers and analysts can discover granular micro-segments (e.g., distinguishing stable high-frequency buyers from seasonal or decaying high-frequency buyers). Keywords Customer Segmentation, Clustering Results, RFM Dataset, K-Means, Time Series Clustering, Dynamic Time Warping, DTW, Business Intelligence, Customer Analytics.