Framework for Spectral Analysis of Resonance Patterns in Quality Factors

We have considered a network topology of several types of weights to describe the dynamical properties of the real environments. In this study, we adopt the red wine quality problem as a case study of it and develop an experimental model using a publicly available dataset. An objective and quantitative approach is needed to assess wine quality. This is because sensory testing, known as tasting, is influenced by the winemaking process, grape origin, barrel type and ageing period, as well as the flavor profile of a particular grape variety. Modelling and spectral analysis is conducted by assigning inter-agent weights to each parameter related to wine quality, using the typical graph of star topologies that have been studied. The obtained resonance patterns of eigenvectors are decoded in this paper. The impact of each parameter on wine quality is analyzed across factors to identify the parameter that improve wine quality. The eigenvalues and eigenvectors obtained from the model are used in this paper to investigate resonance patterns. In our previous work, the largest and smallest eigenvalues have been the main focus of analysis. In this paper, we adopt the analysis of resonance patterns of eigenvectors in order to cross-sectionally evaluate factors contributing to wine quality improvement. We present a framework to identify the fundamental laws governing quality dynamics.