Set Up Cluster Centers and Thresholds for Anomaly Detection in Qdrant

This workflow is part of a series designed to prepare datasets for anomaly detection using Qdrant. It sets up cluster centers and calculates threshold scores necessary for identifying anomalies. The workflow employs two methods: the distance matrix approach, which identifies a representative point in each cluster, and the multimodal embedding model approach, which uses text embeddings to find the most representative image. These methods ensure that the dataset is thoroughly prepared for subseque