Final Projects – ML for Earth & Env. Sciences

Fall 2022

Marc O’Callaghan: Comparison of YOLOv4-tiny and YOLOv7-tiny for instream wood detection

Julia Walker: Comparing Machine Learning Algorithms for Land Cover Classification for Vallon de Nant, Switzerland

Fabien Augsburger: Machine Learning Prediction of Moderate and Extreme Wind Bursts in Europe

Étienne Delaloye: Using a Random Forest Classifier and Logistic Regression for Binary Classification of Water Potability

Axelle Bersier: Exploratory analysis of clustering Swiss population using Self-Organizing Maps

Margaux Hofmann: Predicting Snow Line Elevation and Glacier Activity using Machine Learning

Elias Al Alam: Detection of Oil Spills Using Machine Learning

Doruntina Bekolli: Prediction of Rainfall in Australia Using Logistic Regression and Random Forest Algorithms

Douglas Stumpp: Extracting road networks from satellite images with neural networks to contribute to risk mitigation

Haakon Vikesaa: Comparing Logistic Regression and Random Forest Classifiers for Landslide Hazard Assessment in Vaud, Switzerland

Melinda Femminis: Quantitative estimation of the area affected by the 2022 Jagersfontain tailings dam’s collapse

Emmanuel Emezina: Assessment of the EuroSAT Dataset by applying Deep Learning Techniques

Florent Rouge: Instream wood detection using the YOLOv4 algorithm on aerial images of the Spöl river

Max Henking: Classifying train delays based on environmental variables

Jérémie Fragnière: Clustering Tropical Cyclogenesis Events Based on Environmental Predictors and Oceanic Basins

Christophe Reymond: Instream wood detection using the YOLOv4 object detection algorithm

Faye Perchanok: Predicting proglacial lake discharge using machine learning algorithms

Jonathan Cotasson: Comparing algorithms for storm cluster detection

Janbiro Ntamushobora: Tracking the evolution of gentrification in the neighborhoods of Paris

Baptiste Poffet: Quantifying deforestation in the Amazon forest using machine learning

Thomas Krieger: Predicting extreme events with Random Forest and K-Nearest Neighbors