[2/3] Set up medoids (2 types) for anomaly detection (crops dataset)

In the realm of agricultural data analysis, identifying anomalies in crop datasets can be a daunting task. Traditionally, researchers would manually sift through vast amounts of data to pinpoint outliers, a process fraught with human error and inefficiency. This workflow addresses the manual pain point of anomaly detection by automating the setup of medoids for clustering, thereby eliminating the tedious task of determining cluster centers and threshold scores. By automating this process, researchers can focus on interpreting results rather than getting bogged down with data preparation.

This n8n workflow employs a series of nodes to establish medoids for anomaly detection using two distinct approaches. The workflow begins with a manualTrigger node to initiate the process. Following this, multiple httpRequest nodes are utilized to gather necessary data from various sources. The code node is then employed to calculate distance metrics and identify cluster centers based on the 'distance matrix approach' in the upper branch. Simultaneously, the lower branch utilizes additional httpRequest and set nodes to implement the 'multimodal embedding model approach'. The workflow concludes with further httpRequest nodes to finalize the setup, ensuring that the cluster threshold scores are accurately determined.

This workflow is ideal for agricultural data scientists and research teams focused on crop analysis. For instance, a team working on precision agriculture can use this workflow to quickly identify which crops are underperforming based on historical data. Similarly, agronomists conducting field experiments can benefit by automating the anomaly detection process, allowing them to concentrate on making data-driven decisions rather than spending time on manual calculations.

Getting started with this workflow is straightforward. Users can deploy it directly to n8n via FlowEngine, where it can be customized to fit specific datasets or analysis needs. Once deployed, users can modify the parameters for the distance metrics or embedding models to tailor the anomaly detection process, ensuring that it meets the unique requirements of their research projects.