A multivariate extreme value theory approach to anomaly clustering and visualization

Chiapino, Maël; Clémençon, Stephan; Feuillard, Vincent; Sabourin, Anne

doi:10.1007/s00180-019-00913-y

Cited by 8 publications

(5 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Hence, the findings are focused on these two extreme clusters. Chiapino et al (2020) also suggest focusing on extreme values (extreme value theory [EVT]) to analyze patterns and structure in data. Similarly, EVT is extensively used in risk management research to identify data clusters that are associated with high or low values (Embrechts et al , 1999), in our case the most resilient and the most vulnerable other clusters (please see Figure 1).…”

Section: Resultsmentioning

confidence: 99%

Characteristics of crises resilient organizations in the hospitality and tourism industry

Aydogan

Curiel

Antonovica

et al. 2023

IJCHM

View full text Add to dashboard Cite

Purpose COVID-19, like many previous crises, proved once more that some hospitality and tourism organizations are more crises resilient than others. Despite increasing frequency and magnitude of crises, little is known about the features of crises resilient organizations and mitigation strategies they adopt. If the characteristics of such resiliency are identified, those strengths might be targeted. Hence, the purpose of this study is to identify characteristics of crises resilient organizations by analyzing the interface between different organizational characteristics, recovery strategies they adopted and impacts of COVID-19 on individual hospitality and tourism organizations. Design/methodology/approach A global sample of 202 respondents from 20 countries and four continents, representing different sectors of the hospitality and tourism industry, participated in the survey. Descriptive analysis and cluster analysis were used to rank the items and group hospitality and tourism organizations based on their crises resiliency. Findings Service quality, loyal customers, branding, high paid in capital, domestic market base, hygiene and safety image, information and communication technology adoption, product and market diversification and restructuring debts emerged as major characteristics and strategies of crises resilient organizations. Using cluster analysis, four different groups of organizations were identified. Based on the impacts of COVID-19 on these organizations, Cluster-1 emerged as significantly more crises resilient, whereas Cluster-4 organizations were significantly more vulnerable to crises. Their characteristics and mitigation strategies they adopted were discussed. Research limitations/implications The paper not only identified features of crises resilient organizations and successful mitigation strategies but also measured their impact on various performance indicators. Future studies might use characteristics, mitigation strategies and performance indicators identified in this study. Practical implications Based on the findings, tourism organizations would focus on strengthening characteristics and implementing strategies that make crises resilient organizations. Public bodies and destination management would also set their decision criteria based on these findings to create a more resilient tourism industry. Originality/value This research not only identifies how hospitality and tourism organizations are affected by COVID-19 but also how these impacts change based on different organizational characteristics and strategies. Understanding which organizational characteristics affect the crises vulnerability of hospitality and tourism organizations might inform risk and crises management literature and structural design elements in tourism businesses, hence offer both theoretical and practical implications.

show abstract

Section: Resultsmentioning

confidence: 99%

Characteristics of crises resilient organizations in the hospitality and tourism industry

Aydogan

Curiel

Antonovica

et al. 2023

IJCHM

View full text Add to dashboard Cite

show abstract

“…analyzing the extremes of a high dimensional random vector. Such studies can be divided into the following categories: clustering methods [6,7,39], support identification, [32,33,8,9,53,44], Principal Component Analysis of the angular component of extremes [14,41,20], and graphical models for extremes [36,23,1]; see also [24] and the references therein. Our approach is remotely related to the last category: extremal graphical models.…”

Section: Dimensionality Reduction In Evtmentioning

confidence: 99%

Tail inverse regression for dimension reduction with extreme response

Aghbalou¹,

Portier²,

Sabourin³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

We consider the problem of dimensionality reduction for prediction of a target Y ∈ R to be explained by a covariate vector X ∈ R p , with a particular focus on extreme values of Y which are of particular concern for risk management. The general purpose is to reduce the dimensionality of the statistical problem through an orthogonal projection on a lower dimensional subspace of the covariate space. Inspired by the sliced inverse regression (SIR) methods, we develop a novel framework (TIREX, Tail Inverse Regression for EXtreme response) relying on an appropriate notion of tail conditional independence in order to estimate an extreme sufficient dimension reduction (SDR) space of potentially smaller dimension than that of a classical SDR space. We prove the weak convergence of tail empirical processes involved in the estimation procedure and we illustrate the relevance of the proposed approach on simulated and real world data.

show abstract

“…How to use EVT in a machine learning framework has received increasing attention in the past few years. Learning tasks considered so far include anomaly detection Roberts (1999Roberts ( , 2000; Clifton et al (2011); Goix et al (2016); Thomas et al (2017), anomaly clustering Chiapino et al (2019a), unsupervised learning Goix et al (2015), online learning Carpentier & Valko (2014); Achab et al (2017), dimension reduction and support identification Goix et al (2017); Chiapino & Sabourin (2016); Chiapino et al (2019b). The present paper builds upon the methodological framework proposed by Jalalzai et al (2018) for classification in extreme regions.…”

Section: Introductionmentioning

confidence: 97%

Heavy-tailed Representations, Text Polarity Classification & Data Augmentation

Jalalzai¹,

Colombo²,

Clavel³

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

The dominant approaches to text representation in natural language rely on learning embeddings on massive corpora which have convenient properties such as compositionality and distance preservation. In this paper, we develop a novel method to learn a heavy-tailed embedding with desirable regularity properties regarding the distributional tails, which allows to analyze the points far away from the distribution bulk using the framework of multivariate extreme value theory. In particular, a classifier dedicated to the tails of the proposed embedding is obtained which performance outperforms the baseline. This classifier exhibits a scale invariance property which we leverage by introducing a novel text generation method for label preserving dataset augmentation. Numerical experiments on synthetic and real text data demonstrate the relevance of the proposed framework and confirm that this method generates meaningful sentences with controllable attribute, e.g. positive or negative sentiment.

show abstract

A multivariate extreme value theory approach to anomaly clustering and visualization

Cited by 8 publications

References 34 publications

Characteristics of crises resilient organizations in the hospitality and tourism industry

Characteristics of crises resilient organizations in the hospitality and tourism industry

Tail inverse regression for dimension reduction with extreme response

Heavy-tailed Representations, Text Polarity Classification & Data Augmentation

Contact Info

Product

Resources

About