Regularization in the Process of Developing an Artificial Neural Network

IMANOL, BILBAO and JAVIER, BILBAO (2017) Regularization in the Process of Developing an Artificial Neural Network. In: Fifth International Conference on Advances in Computing, Electronics and Communication - ACEC 2017, 27-28 May, 2017, Rome, Italy.

20170612_053906.pdf - Published Version

Download (796kB) | Preview
Official URL:


The process for classifying a set of data can depend on several variables, which can have a non-very direct relation among them. Using mathematical techniques such as regression is one of the most accepted methods. Moreover, in the last time and when number of data is higher and concepts like deep learning are applied, artificial neural networks (ANN) are taking into account as a method to solve these classification systems. But, when these ANN are used, some problems must be resolved in order to obtain good results. Some of these problems are overfitting and underfitting. In this paper, an approach to the resolution of them by means of regularization is dealt with.

Item Type: Conference or Workshop Item (Paper)
Uncontrolled Keywords: artificial neural networks, linear regression, logistic regression, regularization, overfitting, underfitting
Depositing User: Mr. John Steve
Date Deposited: 15 Mar 2019 11:07
Last Modified: 15 Mar 2019 11:07

Actions (login required)

View Item View Item