"Anomaly Detection Using K-Means Clustering with JavaScript" in Visual Studio Magazine

I wrote an article titled “Anomaly Detection Using K-Means Clustering with JavaScript” in the December 2025 edition of Microsoft Visual Studio Magazine. See https://visualstudiomagazine.com/articles/2025/12/01/anomaly-detection-using-k-means-clustering-with-javascript.aspx.

Anomaly detection is the process of examining a dataset to find data items that are unusual in some way. The terms anomaly detection and outlier detection are often used interchangeably.

There are many different anomaly detection techniques. Each technique has pros and cons. My article explains how to identify anomalous data items using k-means data clustering, implemented with the JavaScript language. The idea is simple: cluster the dataset of interest, then find the data items that are most different from their associated cluster means.

The output of the demo program is:

Begin k-means anomaly detection JavaScript

Loading synthetic (200) data

X data:
[  0] -0.1660   0.4406  -0.9998  -0.3953  -0.7065   0.4840
[  1]  0.0776  -0.1616   0.3704  -0.5911   0.7562   0.1568
[  2] -0.9452   0.3409  -0.1654   0.1174  -0.7192   0.8054
[  3]  0.9365  -0.3732   0.3846   0.7528   0.7892   0.1345
[  4] -0.8299  -0.9219  -0.6603   0.7563  -0.8033   0.7955
. . .

Clustering with k=3
Done

Clustering:
  2  0  1  0  1  2  1  2  1  1  2  2  0  2  1  2  1  0  1  2
  1  2  0  2  1  0  1  2  2  1  2  0  0  0  1  0  2  0  2  0
  1  0  0  1  1  0  2  1  2  1  1  2  2  0  0  2  0  2  0  1
  2  1  1  0  2  0  0  1  1  1  0  0  2  1  0  1  1  1  0  1
  0  1  0  1  0  0  2  2  1  1  0  0  0  0  2  1  1  2  1  2
  2  1  1  1  0  0  0  0  2  2  1  2  0  0  0  0  0  1  0  0
  2  0  0  2  1  0  1  2  2  1  2  1  2  1  1  1  1  0  2  1
  2  2  1  2  1  1  0  0  2  0  2  2  1  0  2  0  1  1  0  0
  1  1  0  0  2  2  0  2  2  0  1  0  0  2  1  2  2  2  1  1
  2  1  0  0  2  1  2  2  1  0  0  2  2  2  1  0  1  0  2  1

Means:
[  0]  0.2662  -0.3617   0.5050   0.0062   0.2880   0.2144
[  1] -0.6035   0.0620  -0.1627   0.0374  -0.1405   0.5631
[  2]  0.4208   0.3543  -0.3674  -0.1722   0.0537   0.2609

WCSS: 240.7398

Anomalies:

cluster  idx   distance     data[idx]
-----------------------------------------
  0     [107]  1.6625 [  0.51  0.43  0.97 -0.44 -0.99  0.28 ]
  1     [136]  1.6207 [ -0.05  0.97  0.89  0.29 -0.69  0.49 ]
  2     [138]  1.6812 [  0.15  0.84 -0.87  0.98 -0.89  0.42 ]

End demo

This means data item [0] is assigned to cluster 2, item [1] is assigned to cluster 0, item [2] is assigned to cluster 1, and so on.

The WCSS (within cluster sum of squares) is a measure of how well the data has been clustered, where smaller values are better.

For cluster 0, the data item that belongs to the cluster that is farthest from the mean of cluster 0 is item [107] = (0.5175, 0.4339, 0.9747, -0.4438, -0.9924, 0.2879). That item is a distance of 1.6625 away from the mean of cluster 0. For cluster 1, the most anomalous item is [136], and for cluster 2, the most anomalous item is [138].

Anomaly detection using k-means clustering is just one of many anomaly detection techniques. Different techniques tend to find different types of anomalies. Other clustering-based anomaly detection techniques includes DBSCAN anomaly detection and self-organizing map anomaly detection. Another category of anomaly detection techniques analyze source data to generate a math model of the structure of the data, and then analyze reconstruction error to find anomalous items. Examples includes neural network autoencoder anomaly detection and quadratic regression anomaly detection.

“Amazing Stories” is a science fiction magazine. It was launched in April 1926 and was the first magazine devoted solely to science fiction. The magazine has been published for 99 years (but with significant interupptions).