Cluster analysis

Cluster analysis

A cluster refers to a group of data points or observations with similar characteristics, patterns, or features. Clustering is a technique used in unsupervised learning where the algorithm aims to identify inherent structures or groups within a dataset without predefined labels. The goal is to group data points together based on their similarities and form clusters.

Steps to perform cluster analysis in Lumenore:

Step 1: After accessing “Do You Know,” select “Cluster.”

Step 2: Choose a Schema and click “Next.”

Note: The schema signifies the dataset for analysis. If absent, create one, ensuring prerequisites (A KPI, Date, and Attribute) are met.

Step 3: Select the following to configure:

• Select clustering metric: These metrics, also known as Key Performance Indicators (KPIs), serve as quantitative measures to assess the effectiveness and quality of clustering algorithms. Select a minimum of two metrics or KPIs to evaluate the clustering analysis comprehensively.
• Select clustering attribute: A feature or variable in a dataset that groups similar data points. When conducting clustering analysis, one needs to select specific attributes, such as Date, Agent, Product, etc., based on which the data distribution should be performed to identify meaningful patterns and structure in the dataset. The clustering attribute is a key factor in determining how data points are grouped and differentiated during the clustering process.
• Select clustering bucket: Clustering buckets involve grouping or categorizing data points into clusters during clustering analysis. Choose an attribute, such as Product Category, Country, or Region, to create distinct clustering plots based on that specific characteristic.
• Number of clusters: The desired or expected count of distinct groups or clusters the algorithm should identify within a dataset. Specify according to your needs.

Click “Next.”

Note: If you wish to apply a filter, a window will appear for creating filters. As shown in the Trend Analysis, establish filters by groups or conditions as needed.

Step 4: Users can tailor the insights narrative, outlining all the variables in crafting the insight. Then, click on “Save.”

Step 5: Name the insight for future access (default suggestion provided) and save it.

Step 6: A new window appears; click “Execute Now” to generate insights.

Upon initiation of execution, the system will undergo four background processes. You can also terminate the execution at any point before its completion.

Here, you can see the five clusters of profit and sales.