Challenges 2020
In this second session, 9 other challenges from the Challenge data website are presented:
- " Predicting gender based on brain rhythm " presented by Valentin Thorey from Dreem. The aim of this challenge is to predict the gender of individuals based on forty intervals of their electrocardiogram.
- " Predicting next-day wind power on the energy market " presented by Olivier Vannier from CNR. The aim of this challenge is to predict the energy production of six wind power plants based on meteorological data.
- " Predicting the response time of a Paris Fire Brigade vehicle " presented by Benjamin Berhault from the public body BSPP. The aim of this challenge is to predict the time between the selection of a brigade vehicle and its arrival on site, based on time measured in the past and vehicle data.
- " Predicting Bitcoin direction from sentiment data " presented by Marc des Ligneris of Napoléon AM. The aim of this challenge is to predict future Bitcoin returns by classifying them into three categories.
- " Classification of songs from ten species of Odontocetes " presented by Hervé Glotin from Université de Toulon. The aim of this challenge is to classify the sound recordings of marine animals into ten categories.
- " Failure prediction in a production line " presented by Eric Manouvrier from Valeo. The aim of this challenge is to predict the failure of factory parts during assembly. This prediction can be based on industrial data, angle values, logical measurements and other data.
- " Segmentation of objects from computer-generated images presented by Eliot Angles from PhotoRoom. The aim of this challenge is to build an algorithm to determine where the central object of an image is located. The solution is a set of pixels.
- " Disaggregating the load curve of a dwelling " presented by Coline Baraize from BCM Energy. Given the overall consumption of a dwelling, the aim of this challenge is to determine which part of this consumption is attributed to which appliance among washing machine, refrigerator, television and kettle.
- " Coverage of an equinox by machine learning " presented by Olivier Croissant from Natixis. The aim of this challenge is to build a model producing an optimal option replication strategy based on the asset's value over time and other market values such as volatility.
