Publications associées à Iris de Fisher

Self-Supervised Bayesian representation learning of acoustic emissions from laser powder bed Fusion process for in-situ monitoring

Christian Leinenbach, Sergey Shevchik, Rafal Wróbel, Marc Leparoux

This study presents a self-supervised Bayesian Neural Network (BNN) framework using air-borne Acoustic Emission (AE) to identify different Laser Powder Bed Fusion (LPBF) process regimes such as Lack of Fusion, conduction mode, and keyhole without ground-tr ...

Elsevier Sci Ltd2023

Bias at a Second Glance: A Deep Dive into Bias for German Educational Peer-Review Data Modeling

Vinitra Swamy, Tanja Christina Käser Jacober, Thiemo Wambsganss

Natural Language Processing (NLP) has become increasingly utilized to provide adaptivity in educational applications. However, recent research has highlighted a variety of biases in pre-trained language models. While existing studies investigate bias in di ...

2022

Lane Detection and Lane Changing Phenomena Identification with Data from a Swarm of Drones

Nikolaos Geroliminis, Emmanouil Barmpounakis

In the era of big data, new transportation-related concepts and methodologies need to be proposed for understanding how congestion propagates. pNEUMA, a unique dataset that was acquired during a first-of-its-kind experiment with a swarm of drones over a de ...

2020

Modeling and Analyzing Inquiry Strategies in Open-Ended Learning Environments

Tanja Christina Käser Jacober

Modeling and predicting student learning in computer-based environments often relies solely on sequences of accuracy data. Previous research suggests that it does not only matter what we learn, but also how we learn. The detection and analysis of learning ...

2020

Two-Tier Mapper, an unbiased topology-based clustering method for enhanced global gene expression analysis

Cathrin Brisken, Kathryn Hess Bellwald, Jacques Rougemont, Rachel Marcone

Motivation: Unbiased clustering methods are needed to analyze growing numbers of complex data sets. Currently available clustering methods often depend on parameters that are set by the user, they lack stability, and are not applicable to small data sets. ...

2019

SemiHD: Semi-Supervised Learning Using Hyperdimensional Computing

Samuel Bosch, Mohsen Imani

In the Internet of Things (IoT), the large volume of data generated by sensors poses significant computational challenges in resource-constrained environments. Most existing machine learning algorithms are unable to train a proper model using a significant ...

IEEE2019

The SDSS-IV eBOSS: emission line galaxy catalogues at z approximate to 0.8 and study of systematic errors in the angular clustering

Jean-Paul Richard Kneib, Johan Comparat, Anand Stéphane Raichoor, Xiaoqing Zhou, Timothée Guy Olivier Delubac

We present two wide-field catalogues of photometrically selected emission line galaxies (ELGs) at z approximate to 0.8 covering about 2800 deg(2)over the south galactic cap. The catalogues were obtained using a Fisher discriminant technique described in a ...

Oxford Univ Press2017

Challenges in Algorithmic Mechanism Design

Paul David Dütting

This thesis addresses three challenges in algorithmic mechanism design, which seeks to devise computationally efficient mechanisms consisting of an outcome rule and a payment rule that implement desirable outcomes in strategic equilibrium. The first challe ...

EPFL2013

Inverse Solutions for Brain Computer Interface

José del Rocio Millán Ruiz, Mohit Kumar Goel

Inverse solution allows to estimate sources that generate a given scalp EEG topography. Recently, it has been used in Brain Computer Interfaces (BCIs) to extract robust features based on the hypothesis that projection onto the source space (high dimensiona ...

EPFL2013

Online BCI with Stable Sources

José del Rocio Millán Ruiz, Ricardo Andres Chavarriaga Lozano, Mohit Kumar Goel

In this paper, we show that the estimated intra-cranial sources using source localization on EEG signals can be used for online Brain Computer Interation (BCI) and the discriminant sources remain stable over days. Classifiers are trained on discriminant so ...

2012