A Bio-inspired Ensemble Model for Food Industry Applications

摘要

This paper presents a soft computing robust solution for the food indus-try field with the aim of analysing the olfactory properties of Spanish dry-cured ham. A novel topology preserving version of the Visualization Induced SOM (Vi-SOM), based on the application of the Weighted Voting Superposition (WeVoS) summarization algorithm, is presented in order to calculate the best possible visu-alization of the internal structure of a datasets. The results obtained by this novel model are compared with the ones obtained by its single version -ViSOM- and ver-sus the well-known SOM and WeVOS-SOM. The results clearly demonstrate how the WeVoS-ViSOM outperforms the rest of models.