[PhD’s corner] Josie Signe : Animal welfare : characterizing the diversity within livestock with data mining methods used on data from dairy herd sensors

Josie Signe is one of the #DigitAg cofunded PhDs

 

 

Animal welfare : characterizing the diversity within livestock with data mining methods used on data from dairy herd sensors

  • Start Date: 1st October 2020
  • University: Université de Rennes 1
  • PhD School: MathSTIC
  • Field(s): Computer Science
  • Doctoral Thesis Advisor: Alexandre Termier, Lacodam, Inria
  • Co-supervisors : Peggy Cellier, Insa Rennes, Véronique Masson, Lacodam, Université de Rennes 1
  • Funding: #DigitAg – Inria
  • #DigitAg: Cofunded thesis – Axis 5: Data Mining, Data Analysis and Knowledge Discovery, Challenge 4: ICT and sustainable animal production

Keywords: Data mining, time series, animal welfare, dairy cows

Abstract: Climate change is characterized in particular by episodes of greater heat during the summer, with sometimes very high temperatures, which can be harmful to the well-being and production of dairy animals. This heat stress is also amplified when the humidity is high, as in the case of stormy episodes. Today’s monitoring tools allow the measurement of real-time data, called time series. It is in particular possible to follow the evolution of the temperature of an animal over long time steps, both during lactation but also over several lactations. These data make it possible, for example, to observe that the temperature response to thermal stress is not the same for each animal. These differences could be explained by the temperature threshold of an animal, its adaptive capacities, its individual characteristics (ingestion, milk production, weight…). Being able to characterize, but also predict, the response of an animal to thermal stress constitutes a real challenge for the well-being of animals and the maintenance of their performance. To characterize the heat stress response in cows, we will use discriminant pattern extraction methods, capable of extracting groups with different heat stress responses. In particular, we focus on algorithms for finding out subgroups and characterizing for example groups of cows at risk. However, subgroup discovery algorithms are not integrated to handle sets of time series. A first task is therefore the construction of a method for discovering subgroups making it possible to process time series. Another important task is to define more precisely how to characterize these subgroups in order to find characteristics specific to dairy cows.

Contact: josie.signe@inria.fr