Marie Chavent’s research while affiliated with National Institute for Research in Computer Science and Control and other places

What is this page?


This page lists works of an author who doesn't have a ResearchGate profile or hasn't added the works to their profile yet. It is automatically generated from public (personal) data to further our legitimate goal of comprehensive and accurate scientific recordkeeping. If you are this author and want this page removed, please let us know.

Publications (117)


ClimLoco1.0: CLimate variable confidence Interval of Multivariate Linear Observational COnstraint
  • Preprint
  • File available

January 2025

·

5 Reads

Valentin Portmann

·

Marie Chavent

·

Projections of future climate are key to society's adaptation and mitigation plans in response to climate change. Numerical climate models provide projections, but the large dispersion between them makes future climate very uncertain. To refine it, approaches called observational constraints (OC) have been developed. They constrain an ensemble of climate projections by some real-world observations. However, there are many difficulties in dealing with the large literature on OC: the methods are diverse, the mathematical formulation and underlying assumptions used are not always clear, and the methods are often limited to the use of the observation of only one variable. To address these challenges, this article proposes a new statistical model called ClimLoco1.0, which stands for "CLimate variable confidence Interval of Multivariate Linear Observational COnstraint". It describes, in a rigorous way, the confidence interval of a projected variable (its best guess associated with an uncertainty at a confidence level) obtained using a multivariate linear OC. The article is built up in increasing complexity by expressing in three different cases, the last one being ClimLoco1.0, the confidence interval of a projected variable: unconstrained, constrained by multiple real-world observations assumed to be noiseless, and constrained by multiple real-world observations assumed to be noisy. ClimLoco1.0 thus accounts for observational noise (instrumental error and climate-internal variability), which is sometimes neglected in the literature but is important as it reduces the impact of the OC. Furthermore, ClimLoco1.0 accounts for uncertainty rigorously by taking into account the quality of the estimators, which depends, for example, on the number of climate models considered. In addition to providing an interpretation of the mathematical results, this article provides graphical interpretations based on synthetic data.

Download

Dynamics of tumor evolution after Gamma Knife radiosurgery for sporadic vestibular schwannoma: Defining volumetric patterns characterizing individual trajectory

September 2024

·

37 Reads

Neuro-Oncology

·

·

Christine Delsanti

·

[...]

·

Background Definition of tumor control and treatment failure after Gammaknife radiosurgery (GKRS) for vestibular schwannoma (VS) is still debated. The lack of knowledge on the dynamics of tumor evolution can lead to misinterpretation and subsequent inappropriate second treatment. The aim of this study was to evaluate the post-GKRS dynamics of evolution of tumor volume, and characterize volumetric patterns. Methods We included patients with sporadic VS treated by GKRS with an MRI follow-up of minimum 3 years. A clustering in 2 steps was performed: definition of the patterns of evolution based on a subset of patients with the most comprehensive follow-up, then assignment of the remaining patients on a best fit basis. The minimum length of follow-up was assessed by measuring the consistency of the clusters over time (Adjusted Rand Index and Normalized Mutual Information). An analysis of the discriminant variables was finally performed. Results 1,607 patients were included (median follow-up: 67 months). Five patterns were defined with one pattern gathering almost all cases of treatment failure. The clustering at 5 years afforded the highest consistency with long-term follow-up. Discriminant variables for clusters were: sex, initial symptoms, delay of diagnosis, Koos grading, fundus invasion, and number of isocenters. Conclusions The definition of these robust distinct patterns is likely to help tremendously the physicians to distinguish tumor control from potential failure. We advocate for no retreatment decision before 5 years post-GKRS. Further investigations are required to decide if the dynamics of evolution can be predicted at GKRS on an individual basis.



Sparse Weighted K-Means for Groups of Mixed-Type Variables

August 2022

·

26 Reads

Assessing the underlying structure of a dataset is often done by training a clustering procedure on the features describing the data. In practice, while the data may be described by a large number of features, only a minority of them may be actually informative with regard to the structure. Furthermore, redundant features may also bias the clustering, whether one speaks of redundancy in the informative or the uninformative features. The present contribution aims at illustrating two sparse clustering methods designed for mixed data (made of numerical and categorical features). The proposed methods summarise redundant features into groups, and select the most relevant groups of features only in the clustering procedure. The performances and the interpretability of the sparse methods are illustrated on a real-life data set.KeywordsSparse clusteringFeature clusteringFeature selectionGroup of features selectionVariable importance




Certain relationships between Animal Performance, Sensory Quality and Nutritional Quality can be generalized between various experiments on animal of similar types

May 2021

·

36 Reads

·

2 Citations

Livestock Science

In the beef sector, one of the major challenges is to early predict carcass and meat quality and to jointly satisfy the multiple expectations of the various stakeholders. Thus, the objective of this study was to determine if the relationships among carcass, nutritional and sensory qualities established previously by Ellies-Oury et al. (2016) might be generalized to different type of animals. The Longissimus thoracis muscles of 32 young Charolais bulls were analyzed in terms of sensory and nutritional quality (lipid content and fatty acid composition). These parameters of interest were linked together and to animal performances by using a clustering of variables. Longissimus thoracis sensory and nutritional qualities appear sometimes antagonistic. Indeed, some “positive” sensory descriptors (juiciness, overall appreciation, overall flavor and overall odor) are negatively related to PUFA proportions. PUFA proportions are positively associated with carcass weight but in the same time with rancid/fish flavors. Moreover, CLA and trans MUFA proportions are positively associated with the “negative” descriptors of greasy feel and residues. To finish, carcass weight and ADG are negatively associated with some “positive” sensory descriptors except tenderness. It can be concluded from this work that these relationships, that were already established in previous works, are robust between experiments. In order to highlight robust and generalizable relationships in different contexts, it is now appropriate to apply this method to a larger database containing different traits and various characteristics of breed, slaughter ages, animal types, fattening practices, ...


Has breed any effect on beef sensory quality?

May 2021

·

122 Reads

·

25 Citations

Livestock Science

A total of 436 young cattle from 15 cattle breeds were reared in as similar conditions as possible to evaluate the impact of breed on sensory quality of beef from longissimus muscle determined by sensory analysis. Two statistical methods for processing the sensory data were compared. The analysis of variance with or without the panelist effect gave similar conclusions indicating that the robustness of the results was not dependent on the method chosen. The 4 meat descriptors (tenderness, juiciness, beef flavor and off-flavor) placed breeds into 5 groups using an unsupervised classification (hierarchical ascending classification). Aberdeen Angus, Highland and Jersey, that have a high lipid content in the muscle studied, differed from the other breeds in that they had a higher beef flavour. The dual-purpose and rustic breeds, Simmental, Casina and Marchigiana, produced significantly less juicy and less tender meat than that from breeds selected for meat production. Overall, despite significant differences previously identified for animal, carcass, muscle and beef traits for the same animals, differences in sensory scores between most of the breeds were small, with only significant differences between the few breeds that had extreme sensory profiles (such as Simmental and Pirenaica).



Figure 2. Representation of the functioning line for the pair of variables (FNIN1, W2AR) on the stabilized point four. The value of each engine is represented by the colors in the orthogonal space estimated with the RPCA.
Figure 3. Shapley plot of the engine 41. 0.40 is the average anomaly score and 0.47 is the predicted score of the engine 41. Feature value with positive φ increase the score from 0.40 to 0.47, and negative value of φ decrease the anomaly score of the engine.
Figure 6. Diagram of the process of production tests validation using Isolation Forest and Shapley values. The statistical methodology helps to highlight specific engines and measures where further analyses are needed.
Figure 7. Groups weights for each value of λ. The vertical line represents the selected clustering solution where three groups of variables have non-zero weights.
Figure 9. Smooth distances between prototypes. The background colors indicate the distances between neighboring prototypes where pink corresponds to larges distances.

+1

Automatic detection of rare observations during production tests using statistical models

November 2020

·

82 Reads

Annual Conference of the PHM Society

Engines are verified through production tests before delivering them to customers. During those tests, lot of measures are taken on different parts of the engine, considering multiple physical parameters. Unexpected measures can be observed. For this very reason, it is important to assess if these unusual observations are statistically significant. However, anomaly detection is a difficult problem in unsupervised learning. The obvious reason is that, unlike supervised classification, there is no ground truth against which we could evaluate results. Therefore, we propose a methodology based on two independent statistical algorithms to double check our results. One approach is the Isolation Forest (IF) model which is specific to anomaly detection and able to handle a large number of variables. The goal of the algorithm is to find rare items, events or observations which raise suspicions by differing significantly from the majority of the data and, at the same time, it discriminates non-informative variables to improve. One main issue of IF is its lack of interpretability. Within this scope, we extend the shapley values, interpretation indicators, to the unsupervised context to interpret the model outputs. The second approach is the Self-Organizing Map (SOM) model which has nice properties for data mining by providing both clustering and visual representation. The performance of the method and its interpretability depends on the chosen subset of variables. In this respect, we first implement a sparse-weighted K-means to reduce the input space, allowing the SOM to give an interpretable discretized representation. We apply the two methodologies on data on aircraft engines measurements. Both approaches show similar results which are easily interpretable and exploitable by the experts.


Citations (61)


... Consequently, when the sound rays reach the same descent depth, their propagation distance significantly increases. A high correlation between research features may cause the OOB error to underestimate key features [59], resulting in the highest feature importance of each change in sound speed structure affecting the convergence zone distance not exceeding 16%. The Shapley values, however, suggest a more reliable conclusion: CD is identified as the key factor influencing the waveguide of the convergence zone on the warm-water side, with an average feature importance of 21.98% across the three surrogate datasets. ...

Reference:

Data-Driven Analysis of Ocean Fronts’ Impact on Acoustic Propagation: Process Understanding and Machine Learning Applications, Focusing on the Kuroshio Extension Front
Handling Correlations in Random Forests: which Impacts on Variable Importance and Model Interpretability?
  • Citing Conference Paper
  • January 2021

... No significant correlation was found between sensory traits from panelists and PUFA content. These observations differ from the findings by Ellies-Oury et al. (2021), which indicated that with longissimus thoracis muscles from young Charolais bulls, sensory traits (juiciness, overall liking and flavor) evaluated by panelists were negatively correlated with PUFA proportions [43]. This may be due to different animal types and production systems, and also different FA units (content or proportion) used in the different studies. ...

Certain relationships between Animal Performance, Sensory Quality and Nutritional Quality can be generalized between various experiments on animal of similar types
  • Citing Article
  • May 2021

Livestock Science

... Il a été parfois observé qu'il existait des différences significatives de tendreté, jutosité ou flaveur entre certaines races (Micol et al., 2010). Néanmoins, une étude récente n'a observé aucune différence significative de qualité sensorielle entre les races quand les animaux sont élevés et la viande maturée dans les mêmes conditions (Conanec et al., 2021), suggérant que les différences observées seraient dues à une interaction avec d'autres facteurs liés aux conduites d'élevage. ...

Has breed any effect on beef sensory quality?
  • Citing Article
  • May 2021

Livestock Science

... Seleksi ini ditujukan untuk peningkatan kualitas genetik pada suatu populasi sapi bali (Garantjang et al., 2020;Warmadewi et al.,2020;Setiaji et al.,2019;Puspitasari et al.,2018;Rahmatullah et al., 2016). Sedangkan kualitas genetik berpengaruh pada sifat-sifat ukuran tubuh pada ternak yang memiliki nilai ekonomi sehingga hal ini membantu evaluasi pemeliharaan dalam upaya peningkatan produktifitas ternak (Li et al., 2022 ;Ellies-Oury et al., 2020). ...

Various Statistical Approaches to Assess and Predict Carcass and Meat Quality Traits

Foods

... Since the 1990′s, the phytoplankton community of Lake Geneva has undergone major changes due to the reoligotrophication of its waters. This re-oligotrophication of the lake favored the growth of species indicative of oligotrophic lakes and resulted in a deepening of phytoplankton activity linked to modifications of the vertical profiles in phosphorus concentration (Anneville et al., 2002(Anneville et al., , 2019. The latter showed extremely low concentrations in the euphotic zone along with a deepening of the phosphocline. ...

The paradox of re‐oligotrophication: the role of bottom–up versus top–down controls on the phytoplankton community
  • Citing Article
  • August 2019

... As an alternative to AMOC, we used AMV variations estimated from three different methods and found indistinguishable results with the AMOCbased IVS (SI- Fig. 6). The lack of evidence for background state dependence of the IVS AMOC-NAO together with the relevance of AMV as a fingerprint of AMOC open some promising perspectives for last millennium paleoclimate studies based on available proxies of the two drivers [73][74][75] , aiming at better investigating multidecadal variations over Europe in presence, this time, of natural, solar and volcano external forcing, as opposed to actual anthropogenic overly dominant influence that has been described here. ...

Reconstructing climatic modes of variability from proxy records using ClimIndRec version 1.0

... In the meat field, an approach using weight aggregation was used by [28] to find a trade-off between nutritional and sensory quality. This study reported that weight setting has a high influence on the final result and must be managed carefully. ...

New Approach Studying Interactions Regarding Trade-Off between Beef Performances and Meat Qualities

Foods

... Our methods refer to applications where a single variable might be of particular interest with respect to its importance for prediction. This must be differentiated from another active field of research on how to use variable importance for variable selection [31][32][33][34][35] possibly based on p-values [36]. Our research has been motivated by an investigation of the role of kidney quality for posttransplant survival and this example can give some guidance on how to apply our method and interpret its results. ...

Statistical model choice including variable selection based on variable importance: A relevant way for biomarkers selection to predict meat tenderness

... Previous large-scale genetic studies of both brain 29,36 and cognitive or behavioral language-related traits 19,20,24 only analyzed common genetic variants (allele frequency in the population ≥1%). Tentative evidence for rare variant associations with right-hemisphere language dominance, involving actin cytoskeleton genes, was found in an exploratory study of 66 unrelated participants 39 . The first exome-wide association studies of the UK Biobank 40,41 included structural brain imaging metrics, but not functional metrics. ...

Genome sequencing for rightward hemispheric language dominance

... ( 1 ∪ 2 ). Here, with only numeric variables, ( ) was measured via the Bravais-Pearson correlation between the variables of the cluster and a numeric synthetic variable summarizing the cluster, such that ( ) = ∑ ² ( , ) = 1 ∈ . is the first principal component of the PCA of the PCAmixdata package in R (Chavent et al., 2022) applied to the data of the cluster (Saracco et al., 2018) ...

Classification de variables et analyse multivariée de données mixtes issues d’une étude BCI

Ingénierie cognitique