Macrowine 2021
IVES 9 IVES Conference Series 9 Beyond classical statistics – data fusion coupled with pattern recognition

Beyond classical statistics – data fusion coupled with pattern recognition

Abstract

AIM: Patterns in data obtained from wine chemical and sensory evaluations are difficult to infer using classical statistics. Pattern recognition can be resolved by coupling data fusion with machine learning techniques, possibly leading to new hypotheses being formed. This study demonstrates the applicability of two pattern recognition approaches using as case study involving Chenin Blanc wines (recently bottled and after two years storage) from young (35 years) vines.

METHODS: Sensory (sorting (Mafata et al. 2020)) and chemical (NMR: nuclear magnetic resonance, HRMS: high resolution mass spectrometry, and UV-Vis: ultraviolet spectrophotometry) data were collected for the young and aged (two years in the bottle) wines. Data sets were combined using multiple factor analysis (MFA). Exploratory unsupervised cluster analysis was performed by agglomerative hierarchical clustering (AHC) and Fuzzy-k means (Bezdek 1981). Optimal cluster conditions were found for both methods and the cophenetic coefficient was used to assess the most confident clustering method.

RESULTS: Since large data sets were fused, the models were very complex. There were no consistent clustering patterns when varying clustering conditions, signalling high similarity between samples. The samples could not confidently be distinguished from one another even at the highest optimized conditions. Although Fuzzy-k means gave more confident clustering, it was still not sufficient for solving classification issues in this sample set.

CONCLUSIONS:

Fuzzy-k means was better at resolving the natural grouping of samples. Coupled to data fusion, it could potentially lead to better pattern recognition, especially for oenological chemical and sensory data. The fuzzy approach should be explored, keeping in mind it is more sensitive to small differences in the data compared to classical statistics.

DOI:

Publication date: September 7, 2021

Issue: Macrowine 2021

Type: Article

Authors

Mpho Mafata, Jeanne

1South African Grape and Wine Research Institute, Department of Viticulture and Oenology, Stellenbosch University & 2School for Data Science and Computational Thinking, Stellenbosch University, South Africa, BRAND, South African Grape and Wine Research Institute, Department of Viticulture and Oenology, Stellenbosch University, South Africa  Astrid, BUICA, South African Grape and Wine Research Institute, Department of Viticulture and Oenology, Stellenbosch University

Contact the author

Keywords

data fusion, pattern recognition, machine learning, artificial intelligence, multiple factor analysis, fuzzy-k means, cluster analysis

Citation

Related articles…

Varietal volatile patterns of Italian white wines

Aroma diversity is one of the most important features in the expression of the varietal and geographic identity and sensory uniqueness of a wine. Italy has one of the largest ampelographic heritages of the world, with more than five hundred different varieties. Among them, many are used for the production of dry still white wines, many classified as Protected Designation of Origins and therefore produced in specific geographical areas with well-defined grape varieties. Chemical and sensory characteristics of the aroma of these wines have never been systematically studied, and the relative diversity has never been described and classified.

Predictive Breeding for Wine Quality: From Sensory Traits to Grapevine Genome

New pathogen resistant varieties allow an efficient and greatly reduced use of fungicides. These new varieties promise, therefore, an enormous potential to reach the European Green Deal aim of a 50% reduction of pesticides in EU agriculture by 2030.

Red wine substituted esters involved in fruity aromatic expression: an enantiomeric approach to understand their sensory impact and their pathway formation

Among red wines ethyl esters, those from short hydroxylated and branched-chain aliphatic acids constitute a family with a particular behavior and sensory importance. They have been previously discussed in the literature [1] and recent studies have established that some of them were strongly involved in of red wines’ fruity aroma [2]. As some among them have an asymmetrical carbon atom, it seemed important to separate their different enantiomers to obtain an accurate assessment of their organoleptic impact. Three chiral esters have been identified, presenting alkyl and/or hydroxyle substituants: ethyl 2-hydroxy-4-methylpentanoate, ethyl 2-methylbutanoate, and ethyl 3-hydroxybutanoate.

Application of treatments to delay the ripening of grape varieties cultivated in valpolicella

Winegrape cultivars are particularly sensitive to temperature and recent changes in climate have advanced the onset of berry ripening, resulting in unbalanced fruit composition at harvest.

Measurement of redox potential as a new analytical winegrowing tool

Excell laboratory has initiated the development of an analytical method based on electrochemistry to evaluate the ability of wines to undergo or resist to oxidative phenomena. Electrochemistry is a powerful tool to probe reactions involving electron transfers and offers possibility of real-time measurements. In that context, the laboratory has implemented electrochemical analysis to assess oxidation state of different wine matrices but also in order to evaluate oxidative or reduced character of leaf and soil. Initially, our laboratory focused on dosage of compounds involved in responses of plant stresses and we were also interested in microbiological activity of soils. These analyses were compared with the measurement of redox potential (Eh) and pH which are two fundamental variables involved in the modulation of plant metabolism. Indeed, the variation of redox states of the plant reflects its biological activity but also its capacity to absorb nutriments. The Eh-pH conditions mainly determine metabolic processes involved in soil and leaf and our goal is to determine if this combined analytical approach will be sufficiently precise to detect biological evolutions (plant health, parasitic attack…).

Macrowine 2021
IVES 9 IVES Conference Series 9 Beyond classical statistics – data fusion coupled with pattern recognition

Beyond classical statistics – data fusion coupled with pattern recognition

Abstract

Content of the article

References

Section for all references

DOI:

Publication date: September 7, 2021

Issue: (ex: Issue: Terclim 2023)

Type: typeofpublication

Authors

author1, author2, author3

Presenting author

Description

List of affiliations ¹ ² ³

Contact the author

Email address (with mailto: link)

Keywords

List of different keywords (keyword1, keyword2, keyword3)

Tags

Citation

Related articles…

Towards adaptation to climate change in Rioja: Quality evaluation of wines obtained from Grenache x Tempranillo selections

The wine sector is of great relevance and tradition in Mediterranean countries, however, it may be most susceptible to climate change. In recent years, wine production is facing changes worldwide, both at environmental as well as commercial levels, due to global warming and the shift in consumers’ preferences. Wine growers and wine makers are in search of solutions that allow to face these new challenges. One of the most promising initiatives in the long term is the introduction of new plant materials, specifically intraspecific hybridizations between premium varieties that may improve traditional germplasm in its adaptation to climate change. These inter-varietal crosses have the potential to generate quality wines, whilst maintaining the regional typicity, and constitute an attractive alternative for the consumer due to their sensory attributes. In this study, we have evaluated wines from 29 intraspecific Garnacha x Tempranillo hybrids in two different locations, with the aim to assess their oenological potential and sensory attributes. Thirteen of the selections were white and 16 were red. Microvinifications were conducted with two or three replications depending on grape availability. Conventional oenological parameters were determined for all wines. The sensory evaluation and hedonic scores were given by five experts. Red selections obtained higher quality scores than white ones. Among the white selections with higher quality scores, GT-41 Varea and GT-159 Varea outstand, due to their high total acidity and high malic acid content. Regarding red selections, GT-57 Varea and GT-57 UR were perceived as higher in quality, highlighted for their moderate alcoholic and high anthocyanin content. Our results indicate that intraspecific hybridization may be a powerful tool for adapting traditional cultivars to climate change in Rioja.

Grapevine vigour is correlated with N-mineralization potential of soil from selected cool climate vineyards in Victoria, Australia

Excess vigour has been a problem on fertile soils under high rainfall in many cool climate regions of Australia. High and low vigour blocks were selected in vineyards of the cool climate regions of King Valley, Yarra Valley and Mornington Peninsula, Victoria.

Anthocyanin accumulation and extractability during the maturation of the grapes of three varieties

Anthocyanin accumulation and extractability were studied in Tannat, Cabernet Sauvignon and Merlot grapes produced in the south of Uruguay in two consecutive seasons. Typical cultivation situations employed in the region for each variety were considered. A follow-up was carried out, considering 60 plants per vineyard, and the harvest was determined according to the technological indices of maturity. Samples of grapes were taken in duplicate in each vineyard periodically along grape maturation. The basic composition, polyphenolic potential and anthocyanin extractability were determined. Also, half of grapes were frozen and later peeled; skin extractions over 24 hs with a solution of 12% ethanol and pH 3.2 were carried out. The anthocyanin contents of the extracts obtained were determined by HPLC-DAD. The levels of anthocyanins reached the highest values before technological maturity. Anthocyanin extractability had a decrease during grape maturation.

Evaluation of the site index model for viticultural zoning

Une variable composite, dénommée Indice de Site (SI), intégrant les propriétés physiques du sol et le mésoclimat, avait été proposée pour caractériser les terroirs dans le cadre d’une étude des vignobles de Cabernet Sauvignon de Hawke’s Bay en Nouvelle Zélande.

Response of Shiraz/101‐14 mgt to in‐row vine spacing

Knowledge of vine reaction to plant spacing under high potential soil conditions is restricted. This study was done to determine effects of vine spacing