Which common point is exist between Feature Selection, Decision Making and Reliability?
Feature selection is one of the steps in decision making using machine learning and artificial intelligence algorithms and reliability is a feature you add to your system when you make it fault tolerant, flexible, minimum error rate, ...etc. of features that constitute reliability concept.
Regards
This is a very interesting question. I think we find ourselves in an era asking us to extract as much information as possible from the huge amount of data lying around repositories or in databases. From an image processing viewpoint we (as scientific community) have made much of progress so far towards the information extraction step at the expense of a pondered choice of time consuming training in artificial intelligence, considering the high level of computational complexity of current machine learning and deep learning systems.
We still need some progress in terms of trade-off between the overall system performance and the training phase. For this purpose, steps such as feature selection and parameter tuning inside Artificial Intelligence field play a critical and fundamental role as well.
The common point between the steps you mentioned in your question depends on the abstractness level of your topic. What I mean is that all depends on the application domain you want to investigate. The wider the application domain, the more "general purpose" the feature description you need...
In my opinion, the most important point linking all of these is the 'quality and quantity' of features. This ensures that you have make accurate 'decisions' which makjes your system 'reliable'.
Poor quality features- bad decision making
Too many features- excess computational time.
That's why right quality and quantity of features is a must!
Best regards!
- Badges
- Science topic
- Similar topics
- Computer Science