A5031505076- Solar Radiation and Photovoltaics
- Building Energy and Comfort Optimization
- Energy Load and Power Forecasting
- Impact of Light on Environment and Health
- Land Use and Ecosystem Services
- Wind and Air Flow Studies
- Urban Heat Island Mitigation
- Frailty in Older Adults
- Nutrition and Health in Aging
- Urban Transport and Accessibility
- Photovoltaic System Optimization Techniques
- Chronic Disease Management Strategies
- Geothermal Energy Systems and Applications
- Health Systems, Economic Evaluations, Quality of Life
- Medical Coding and Health Information
- Climate change and permafrost
- Urban Design and Spatial Analysis
- Solar Thermal and Photovoltaic Systems
- Statistical Methods in Epidemiology
- Model Reduction and Neural Networks
- Hydrological Forecasting Using AI
- Machine Learning and Data Classification
- Geophysical and Geoelectrical Methods
- Neural Networks and Applications
- Landslides and related hazards
École Polytechnique Fédérale de Lausanne
2016-2023
Université de Montréal
2023
Mila - Quebec Artificial Intelligence Institute
2023
Centre universitaire de médecine générale et santé publique, Lausanne
2021-2022
University of Lausanne
2021-2022
Most claims-based frailty instruments have been designed for group stratification of older populations according to the risk adverse health outcomes and not itself. We aimed develop validate a tool based on one-year hospital discharge data Fried's phenotype (FP).We used three-stage development/validation approach. First, we created clinical knowledge-driven electronic score (eFS) calculated as number deficient organs/systems among 18 critical ones identified from International Statistical...
The very shallow geothermal potential (vSGP) is increasingly recognized as a viable resource for providing clean thermal energy in urban and rural areas. This primarily due to its reliability, low-cost installation, easy maintenance, little constraints regarding ground-related laws policies. We propose methodology extract the theoretical vSGP (installed uppermost 10 m of ground, mostly at depths 1–2 m) national scale Switzerland, based on combination Geographic Information Systems,...
Large scale solar Photovoltaic (PV) deployment on existing building rooftops has proven to be one of the most efficient and viable sources renewable energy in urban areas. As it usually requires a potential analysis over area interest, crucial step is estimate geometric characteristics rooftops. In this paper, we introduce multi-layer machine learning methodology classify 6 roof types, 9 aspect (azimuth) classes 5 slope (tilt) for all Switzerland, using GIS processing. We train Random...
Although the trend of progressing morbidity is widely recognized, there are numerous challenges when studying multimorbidity and patient complexity. For multimorbid or complex patients, prone to fragmented care high health use, novel estimation approaches need be developed.This study aims investigate complexity Swiss residents aged ≥50 years using clustering methodology in claims data.We adopted a based on random forests used 34 pharmacy-based cost groups as only input feature for procedure....
Regional-scale urban residential densification provides an opportunity to tackle multiple challenges of sustainability in cities. But framework for detailed large-scale analysis potentials and their integration with natural capital assess the housing capacity is lacking. Using a combination Machine Learning Random Forests algorithm exploratory data (EDA), we propose density scenarios housing-capacity estimates potential lands Oxford–Cambridge Arc region (whose current population 3.7 million...
Solar power harbors immense potential in mitigating climate change by substantially reducing CO$_{2}$ emissions. Nonetheless, the inherent variability of solar irradiance poses a significant challenge for seamlessly integrating into electrical grid. While majority prior research has centered on employing purely time series-based methodologies forecasting, only limited number studies have taken account factors such as cloud cover or surrounding physical context. In this paper, we put forth...
Solar photovoltaic (PV) deployment on existing building roof-tops has proven to be one of the most viable large scale resources sustainable energy for urban areas. While there have been many studies roof-top integrated PV systems at and neighborhood scale, estimating potential through available roof surface area in however, remains a challenge. This study proposes methodology estimate solar buildings commune level (the smallest administrative division) Switzerland. In addition, while several...
Abstract Clean, safe, affordable and available in the long-term, wind is one of most promising sources renewable energy. Its optimized profitable use, however, requires an estimation potential locations interest, given its very volatile behavior various settings. In present study, we propose a methodology using combination Machine Learning (Random Forests), Geographic Information Systems parametric models to estimate large-scale theoretical speed rural areas over entire Switzerland. The...
The computational complexity of classical numerical methods for solving Partial Differential Equations (PDE) scales significantly as the resolution increases. As an important example, climate predictions require fine spatio-temporal resolutions to resolve all turbulent in fluid simulations. This makes task accurately resolving these computationally out reach even with modern supercomputers. a result, current modelers solve PDEs on grids that are too coarse (3km 200km each side), which...
To assess the PV-potential on rooftops for decentralized electricity supply in Switzerland, potential is compared with following urban-size parameters: (a) urban population, (b) number of buildings, (c) cumulative building ground-floor area, and (d) street length. The coefficients determination R2 between PV these factors are: 0.85, 0.86, 0.95, 0.93. resulting relations show that 1% increases a-d are associated PV-potential: 0.82%, 0.93%, 0.94%, 1%, indicating sublinear a-c, but a linear relation d.
We propose a methodology combining physical modelling and machine learning (ML) to estimate the apparent ground thermal diffusivity at scale of country. Based on temperature time series different depths, we 49 Swiss stations using Fourier analysis. Using geology database, estimations are cross-validated with typical values for common rocks. Random Forests, an ML algorithm, used train model previous as output multiple geological, elevation features. The model, showing testing error 16.5%, is...
We propose a novel framework to address the problem of detecting anomalies in building electricity consumption profiles.Our method is based on two sequential steps, which combine machine learning clustering and regression methods.The first step separates weekly anomalous profiles from regular ones, for selected timespan.This achieved through an unsupervised applied representation two-dimensional space.The results are used train model predicts future behavior time series.Any measured deviates...