Anomaly detection using Gaussian vs isolation forest/SVM

Hi, The video covers anomlay dection using Gaussian. Compared to other algo like isolation forest and SVM, is this method better? And how can we better decide which algo/technique to use?

Hello @Joy_Ong,

On the one hand, I think you can research the limitations of each approach, for example, for our gaussian-based anomaly detection algorithm, you probably want to work with continuous features that are gaussian distributed and that the features are independent of each other. Then you can judge whether your dataset is challenging the limitations or not.

On the other hand, if more than one approaches is on the table, you might evaluate their performance over a validation dataset and pick the better performed one.


1 Like