Subhasish Basak

Research & Project experiences : 2022

2022 Projects

Find more projects on my GitHub Page!

Find more on GitHub

GdR MASCOT-NUM 2022

Sigma Clermont, Clermont Ferrand, France

7-9 June 2022

Poster presentation at the annual meeting of the MASCOT-NUM research network at Clermont Ferrand, France. Titled "Bayesian multi-objective optimization for quantitative risk assessment in microbiology", a joint work with Julien Bect (L2S), Laurent Guillier (ANSES), Fanny Tenenhaus-Aziza (CNIEL), Janushan Christy (ACTALIA) and Emmanuel Vazquez (L2S).

This work is part of the ArtiSaneFood project (grant number : ANR-18-PRIM-0015) which is part of the PRIMA program supported by the European Union.

Keywords : ArtisaneFood, Bayesian optimization

53èmes Journées de Statistique

Université Claude Bernard, Lyon, France

13-17 June, 2022

Our article "Integration of bounded monotone functions: Revisiting the nonsequential case, with a focus on unbiased Monte Carlo (randomized) methods", with Julien Bect (L2S) and Emmanuel vazquez (L2S), got accepted at the 53èmes Journées de Statistique.

Dans cet article, nous revisitons le problème de l’intégration numérique d’une fonction monotone bornée, en nous concentrant sur la classe des méthodes de Monte Carlo non séquentielles.

Keywords : Monotone integration, stratified sampling

ONE – Conference 2022

Square, Bruxelles, Belgique

20 June 2022

As a part of the ONE – Health, Environment, Society – Conference attended a workshop for One Health professionals on efficient exchange of mathematical models and data analysis procedures with the harmonized FSKX format.

The objective of the workshop is to introduce the Food Safety Knowledge eXchange (FSKX) format and to promote its adoption by One Health professionals and modellers. FSKX supports the One Health community in its efforts to share and re-use mathematical models, as well as data analysis procedures. Efficient exchange of such knowledge across different One Health sectors (e.g. in risk assessment, hazard identification and control) is a key ingredient for transparent, multi-actor decision-making.

Keywords : Quantitative Risk Assessment (QRA), Microbiology, FSKX, RAKIP

ArtisaneFood annual meeting 2022

Bologne, Italie

7-9 July 2022

As a part of the French unite of the project ArtisaneFood, with Fanny Tenenhaus-Aziza (CNIEL) we presented our ongoing work on Quantitative risk assessment of HUS in french raw milk soft cheese. The presentation was made to the representatives of the other participating countries in the project, including Italy, Portugal, Spain, Maroc, Greece and Tunisia.

The results from the project implementation were shared among all partners in half-hour presentations, and decided upon the actions to be taken to tackle delays due to COVID-19 outbreak were decided by the consortium.

Keywords: Quantitative Risk Assessment (QRA), Microbiology, ArtisaneFood, Bayesian optimization

IDESSAI Summer school

MPI-SWS, Saarbrucken, Germany

29 August - 2 September 2022

Poster presentation at the 2nd Inria-DFKI European Summer School on AI, organized by the two renowned German and French AI institutes, DFKI and Inria.

Abstract : As a part of the European project ArtiSaneFood, the primary goal of this collaborative work between ANSES, CNIEL and L2S is to establish efficient bio-intervention strategies for cheese producers in France, to “economically” reduce the risk of Haemolytic Uremic Syndrome (HUS) caused by Shiga-Toxin producing Escherichia Coli (STEC) present in raw-milk soft cheese. This translates into a multiobjective optimization problem of a stochastic simulator based on a quantitative risk assessment (QRA) model proposed by Perrin et al. (2014), to estimate the Pareto optimal solutions for the process intervention parameters.

Keywords : Pareto Active Learning, Quantitative Risk Assessment (QRA), Bayesian optimization

IDF World Dairy Summit 2022

New Delhi, India

12 - 15 September 2022

Poster presentation at the International Dairy Federation - World Dairy Summit as a part of the French delegation with Dihia Aguenihanai (CNIEL, INRAE) and Thierry Geslain (CNIEL).

The IDF World Dairy Summit is an annual meeting of the global dairy sector, bringing together approximately 1500 participants from all over the world. The participant profile includes CEOs and employees of dairy processing companies, dairy farmers, suppliers to the dairy industry, academicians, government representatives, etc. The summit is composed of a series of scientific and technical conferences and social events including a Welcome Reception, Farmers' Dinner, Gala Dinner as well as Technical and Social Tours.

Keywords : Pareto Active Learning, Quantitative Risk Assessment (QRA), Bayesian optimization

Journée de doctorats, L2S

CentraleSupélec, Université Paris-Saclay

15 September 2022

This was the annual PhD students day (Journée de doctorats) at the Laboratoire de signaux et systemes (L2S), CentraleSupelec, Université Paris-Saclay. All the second year students (in frame Lisa Poirier--Herbeck, CEA) made a presentation of their ongoing work.

Abstract : Bayesian optimization for stochastic simulator using Pareto Active Learning, PALS is easy to implement and an efficient algorithm that uses Gaussian process regression to build surrogate models for the black box expensive simulator to be optimized. It uses confidence rectangles to classify the points in the input design and estimates the Pareto set.

Keywords : Bayesian optimization, Pareto Active Learning

COPIL ArtisaneFood 2022

Caen, Normandie

16 September 2022

The Comité de Pilotage for the ArtisaneFood project at the French national level was organised at Caen, Normandie in collaboration with CNIEL, ANSES, L2S - UPSay, ACTALIA, Fromage AOP de Normandie and producers of french raw milk soft cheese.

Avec l’aide de l’ODG Camembert de Normandie et également Eloïse Modic, la adjointe de l'Association de Gestion des ODG Laitiers Normands, nous avons organiser une visite de laiterie fabricant du Camembert de Normandie. Merci aussi à Coopérative Isigny-Sainte Mère pour l'accueil.

Keywords/Technologies Used: Camembert de Normandie, Isigny-Sainte Mère, ArtisaneFood

Journée de doctorats, ANSES

Maison Alfort, France

18-19 October 2022

This was the annual PhD students day (Journée de doctorats) at Agence Nationale de Securité et Sanitaire (ANSES).

Abstract : We consider the problem of "economically" minimizing the risk HUS form the consumption of raw milk soft cheese. We use a Quantittive Risk Assessment (QRA) model based on the work by Perrin et al (2014), to estimate the risk. Bayesian optimization algorithms are used for optimizing this stochastic simulator using Pareto Active Learning.

Keywords : Bayesian optimization, Pareto Active Learning