STAGE Junior Data Scientist H/F

Title : The impact of extreme climate events on food crops yields predictions using deep neural networks.
Topic : Deep learning, Machine learning, satellite images, meteorological & oceanographic data, agriculture

General presentation of the topic :
Founded in 2013, QuantCube Technology is a leading start-up specialized in Artificial Intelligence applied to macroeconomics and financial predictions. It is composed of a team of 30+ people, including 15+ data scientists in different fields NLP, computer vision and graph analytics.
The company has created the next generation of financial and economic indicators - called Smart Data - by crossing and analysing multiple data sources such as social media, satellite or drone pictures, telecoms data, oceanographic and meteorological data, etc.
Its customers include international organizations, economics institutions, investment banks and asset managers.
From multi-lingual NLP for text analytics to Deep Learning for imagery analytics, QuantCube Technology is also involved in data science competitions and in academics research.

Objective of the internship
QuantCube Technology has set up a unique data warehouse hosting billions of heterogeneous and alternative data from various sources, which provide excellent opportunities for data mining and calibrating any AI models.
Therefore, the aim of the R&D team is to be at the forefront of the state-of-the art of algorithms associated with high-performing computing in the computer vision field.
Within the computer vision team, the objective of this internship is to add the impact of extreme climatographic events, such as El Nino, global warming, extreme rainfalls and extreme draughts to the algorithms making food crops yields predictions. The introduction of non-local spatial and temporal information could be added by modifying the structure of the deep neural networks used for food crops yields predictions or by ad-hoc modification of the oceanographic & meteorological features already included in the model.
During this internship the student will have to work on the state-of-the art algorithms for climatic extreme events'predictions, based on multiple data sources, and food crops yields predictions using satellite images as well (Deep Neural Networks). For this purpose, the student will adapt, develop and implement machine-learning algorithms to fill those objectives.