Modeling a second-generation glucose oxidase biosensor with statistical machine learning methods

Livier Renteria-Gutierrez, Lluis A. Belanche-Muñoz, Felix F. Gonzalez-Navarro, Margarita Stilianova-Stoytcheva


The biosensor is an analytical compact device that embodies a biological detection element and a chemical transducer that converts the biological signal into an electrical output that is proportional to the concentration of the com pund that is desired to assess. The biosensors are high-technology monitoring tools that offer electrochemical measures in a continuous, fast and highly sensitive fashion. Their potential use in real-life applications span from medical and health tasks to environmental monitoring problems. This chapter focuses particularly on Glucose Oxidase Amperometric Biosensors (GOAB) modeling, in order to understand their dynamics and behavior by automated computerized algorithms. In this contribution, we step aside from traditional modeling techniques –mathematical equations and simulation models– by using statistical learning methods. We model the continuous amperometric response of a GOAB by means of several classic and statistical learning methods. Specifically, kernel-based regression techniques are used, such as the support vector machines, one of the best machine learning technique nowadays. We report promising experimental results on a GOAB data set, which is modeled by a non-linear regression method allowing a very low prediction error using a rather simple model of the biosensor output.

Palabras clave

biosensor, glucose oxidase, learning method

Texto completo: PDF


  • No hay Refbacks actualmente.

OmniaScience, 2011-2018 -