In AI based fault classification methods how can i justify a particular feature extraction method (like frequency filtering method) over other methods and how the use of a particular feature selection method (in my case i used Binary Particle Swarm Optimization, BPSO) over other methods like PCA, ReliefF etc?
Joseph, as I understood from your question you want to know which feature extraction is better to use with AI-based fault classification is that right?
If that is the case you cannot until you get the final results since AI is a black box. In my opinion and As J. Rafiee
said you have to test all possible different feature extraction methods.
Feature extraction methods are used to represent a set of samples with a finite number of distinct or continuous values/properties, which are then used to represent the samples (hoping that they will help the decision-making process). For fault classification, frequency analysis, domain transformation methods, and statistical features usually reduce a great deal of noise and make the samples more classifiable.
The sole purpose of feature selection is to reduce the amount of data/features to decrease time and space complexities, without increasing the loss/ decreasing the accuracy too much. For fault classification, genetic algorithms and PSO work quite well.
In both cases, the employed technique will significantly affect the result. Therefore, trying out multiple techniques is imperative.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Artificial Intelligence