Laboratory: IMT Nord Europe, CERI Systèmes Numériques, Lille, France
Duration: 5 months
Supervisors: Christelle Garnier and Anne Savard
Context:
Air pollution monitoring, along with various environmental time-series measurements such as heat waves, rainfall or snowfall, noise and humidity, plays critical role in evaluating trends and impact on human well-being worldwide. In particular, exposure to particulate matter (PM), one of the most dangerous air pollutants, is linked to worsening of respiratory and cardiovascular diseases that can lead to premature death.
In recent years, the development of low-cost pollution sensors has enhanced environment monitoring strategies by improving spatial coverage. Their cost, size and ease of use enable to deploy large networks, generating large datasets that require robust characterization and analysis methods.
Anomaly detection, that consists in identifying data that deviate from their expected behavior, is essential in such networks.
Database consolidation is hence critical to ensure that stored values are validated, correct, and reliable, enabling informed decision-making by communities and institutions. Given the repetitive and time-consuming nature of this process, automating this task is a clear priority.
Objectives:
Within this internship/PFE, we intent to focus on the following objectives:
1) Dataset consolidation: This objective focuses on enhancing the data quality of our datasets by correcting potential biases, filling in missing values and removing known anomalies pollution events.
2) Time series forecasting: This objective involves implementing and comparing several time series prediction approaches, including pattern-based prediction, e.g. Facebook Prophet, and graph-based models, building on our previous works [1, 2]. We will focus on the tradeoff forecasting accuracy vs. computational complexity of the solution.
How to apply:
The applicants should have a strong background in either data science or applied mathematics. Interested candidates have to send their detailed CV, academic records, at least one academic referee and a short motivation letter via email to the contacts below.
Applications will be considered as they arrive, therefore early application is highly encouraged.
Contacts:
Anne Savard
Email: anne.savard@imt-nord-europe.fr
Webpage: https://recherche.imt-nord-europe.fr/personnel/savard-anne/
Christelle Garnier
Email: christelle.garnier@imt-nord-europe.fr
Webpage: https://recherche.imt-nord-europe.fr/personnel/garnier-christelle/
References:
[1] S. Masmoudi, C. Garnier, A. Savard, V. Itier, S. Sauvage, F. Bulot, P. Kaluzny, « Enhanced sensor environment graph based deep learning approach for air quality anomaly detection », EUSIPCO, Lyon, France, 2024
[2] S. Masmoudi, A. Savard, C. Garnier, V. Itier, P. Kaluzny, S. Sauvage, « Graph-based air quality anomaly detection: seasonality effects », to be submitted to IEEE Sensors, 2025
