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…

Know thy enemy: oxygen or storage temperature?

It is well known that high oxygen levels and high ageing temperatures are detrimental to white wine’s composition and ageing capacity. However, these results, though valuable

Carbohydrate dynamics in Shiraz to determine seasonal allocation to the perennial and annual parts in respect to climatic challenges

The dynamic changes of non-structural carbohydrates (NSC) in grapevines during the growing season is driven by phenological events and environmental factors.

Estimation of plant hydraulics of grapevine in various «terroirs» in the Canton of Vaud (Switzerland)

The study of the physiological behaviour of the grapevine (cv. Chasselas), and of plant hydraulics in particular, was conducted on various « terroirs » in the Canton of Vaud (Switzerland) between 2001 and 2003 by Agroscope Changins-Wädenswil ACW, in collaboration with the firm I. Letessier (SIGALES) in Grenoble and the Federal Polytechnic School of Lausanne (EPFL). An evaluation of the vine plant hydraulics was made by means of physiological indicators (leaf and stem water potentials, transpiration and leaf stomatal conductance, carbon isotope discrimination and a model of transpirable soil water), in relation to estimations of the soil water reservoir and climatic factors.

Wine without added SO2: Oxygen impact and color evolution during red wine aging

SO2 play a major role in wine stability and evolution during its aging and storage. Winemaking without SO2 is a big challenge for the winemakers since the lack of SO2 affects directly the wine chemical evolution such as the aromas compounds as well as the phenolic compounds. During the red wine aging, phenolic compounds such as anthocyanin, responsible of the red wine colour, and tannins, responsible of the mouthfeel organoleptic properties of wine, evolved quickly from the winemaking process to aging [1]. A lot of new interaction and molecules occurred lead by oxygen[2] thus the lack of SO2 will induce wine properties changes. Nowadays, the phenolic composition of the wine without added SO2 have not been clearly reported.

VOLATILE COMPOUNDS AND SENSORY PROFILE OF NEBBIOLO RED WINES TREATED WITH WOOD FORMATS ALTERNATIVE TO BARRELS

In winemaking, the use of wood products alternative to barrels, has become a useful tool for the achievement of numerous oenological objectives, including the fast release of desirable volatile and polyphenolic compounds, colour stabilization, and important economic advantages if compared to the traditional barrel production. Among a huge array of variables, the wood format, the vinification protocol, especially the moment of the infusion of the woods and the exposed surface area of the alternative woods are of relevant significance, since they may influence the speed and intensity of the aroma transfer from the wood to the wine defining different sensory profiles.

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…

Frost risk projections in a changing climate are highly sensitive in time and space to frost modelling approaches

Late spring frost is a major challenge for various winegrowing regions across the world, its occurrence often leading to important yield losses and/or plant failure. Despite a significant increase in minimum temperatures worldwide, the spatial and temporal evolution of spring frost risk under a warmer climate remains largely uncertain. Recent projections of spring frost risk for viticulture in Europe throughout the 21st century show that its evolution strongly depends on the model approach used to simulate budburst. Furthermore, the frost damage modelling methods used in these projections are usually not assessed through comparison to field observations and/or frost damage reports.
The present study aims at comparing frost risk projections simulated using six spring frost models based on two approaches: a) models considering a fixed damage threshold after the predicted budburst date (e.g BRIN, Smoothed-Utah, Growing Degree Days, Fenovitis) and b) models considering a dynamic frost sensitivity threshold based on the predicted grapevine winter/spring dehardening process (e.g. Ferguson model). The capability of each model to simulate an actual frost event for the Vitis vinifera cv. Chadonnay B was previously assessed by comparing simulated cold thermal stress to reports of events with frost damage in Chablis, the northernmost winegrowing region of Burgundy. Models exhibited scores of κ > 0.65 when reproducing the frost/non-frost damage years and an accuracy ranging from 0.82 to 0.90.
Spring frost risk projections throughout the 21st century were performed for all winegrowing subregions of Bourgogne-Franche-Comté under two CMIP5 concentration pathways (4.5 and 8.5) using statistically downscaled 8×8 km daily air temperature and humidity of 13 climate models. Contrasting results with region-specific spring frost risk trends were observed. Three out of five models show a decrease in the frequency of frost years across the whole study area while the other two show an increase that is more or less pronounced depending on winegrowing subregion. Our findings indicate that the lack of accuracy in grapevine budburst and dehardening models makes climate projections of spring frost risk highly uncertain for grapevine cultivation regions.

NEAR INFRARED SPECTROSCOPY FOR THE ESTIMATION OF TEMPRANILLO BLANCO VOLATILE COMPOSITION ALONG GRAPE MATURATION

Grape volatile compounds are mainly responsible for wine aroma, so it is important to know the va-rietal aromatic composition throughout ripening process. Currently, there are no tools that allow mea-suring the aromatic composition of grapes, in intact berries and periodically, throughout ripening, in the vineyard or in the winery. For this reason, this work evaluated the use of near infrared spectroscopy (NIR) to estimate the aromatic composition and total soluble solids (TSS) of Tempranillo Blanco berries during ripening. For this purpose, NIR spectra (1100-2100 nm) were acquired from 240 samples of in-tact berries, collected at different dates, from veraison to overripening.

Validation of a high-throughput method for the quantification of volatile carbonyl compounds in wine and its use in accelerated ageing experiments

the aim of this study was the optimization and validation of a robust and comprehensive method for the determination of volatile carbonyl compounds (VCCs) in wines

On quality assurance of winemaking components

This report examines product quality assurance issues arising when technological aids and food additives are utilized in winemaking.

Comparative study of the volatile substances and ellagitannins released to wine by barrels of Quercus pyrenaica, Quercus petraea and Quercus alba

The aim of the study was to study the volatile substances and ellagitannins released to wine by barrels of Quercus pyrenaica (Spanish Oak) in comparison with barrels of Quercus petraea (French Oak) and Quercus alba (American Oak) as well as to determine their sensory impact.