The Minimum Classification Error (MCE) algorithm is a variant of the LVQ method, namely its version LVQ2.1. It tries to ensure the fastest convergence of the solution based on Hebbian learning that updates the winning prototypes by competition.
The difference of the MCE method is that not only the nearest neighbor of each sample in the training set is searched, but also the nearest neighbor with different output and the one with the same output.
In the event that the sample falls within an established window between the two, the one with a different output will be moved away and the one with the same output will be moved closer, in order to improve the classification results.
This is an evolution of the LVQ
Reference: Biing-Hwang Juang and Shigeru Katagiri. Discriminative learning for minimum error classication.Reference.
An article published in April 2021 by Óscar Jiménez El Confidencial, was titled "34,000 M prize for banks for applying well i [...]
Read More »Credit scoring is a system used to rate credits and thus try to automate the decision making process at the time of purchasing a loan, and to [...]
Read More »Artificial intelligence (AI) can change the way sales channels and customers are managed for manufacturers and distributors of consumer products, and can [...]
Read More »When seeking financing for companies, one of the most widely used formulas today is factoring. This is a resource that is not always [....]
Read More »
