Popular Post


Tuesday, December 20, 2022

SETDS (Statistical Exception and Trend Detection System) vs. AWS/SETDS (Split, Extract, Test, Decide, and Score). Is OpenAI right?

 My SETDS (Statistical Exception and Trend Detection System) method was created about 20 years ago. 

Playing with OpenAI https://chat.openai.com/chat I was giving the following information:

"SETDS (Split, Extract, Test, Decide, and Score) is a method for anomaly detection that was developed by researchers at Amazon Web Services (AWS). The method involves the following steps:

  1. Split: The data is split into a training set and a testing set.

  2. Extract: Features are extracted from the training set.

  3. Test: A model is trained on the training set and tested on the testing set.

  4. Decide: Anomaly scores are calculated for each data point in the testing set based on the model's performance.

  5. Score: The anomaly scores are used to identify data points that are likely to be anomalous.

The SETDS method is designed to be fast and efficient, and it has been used in a number of applications, including detecting anomalies in log data and in real-time streaming data. It is one of many methods that can be used for anomaly detection, and the suitability of the method will depend on the specific characteristics of the data and the goals of the analysis."


Interesting that my SETDS also does some kind of scoring by calculating EV - Exception Value. 

Also interesting that I was not able to verify by Google search the information AI bot chat provided....

No comments:

Post a Comment