Name: Edouard Fouché
Type: User
Bio: Data Scientist, Postdoc @ KIT. Especially interested in Data Science, AI, Python, Scala. Hobbyist photographer, Roller Derby Skater
Location: Black Forest, Germany
Blog: http://edouardfouche.com
Edouard Fouché's Projects
Some basic visualization of ArXiv Metadata, see my blog post "arXiv and the Information Explosion"
AutoPhrase: Automated Phrase Mining from Massive Text Corpora
A Scala project that creates solutions for the knight's tour problem (closed and/or complete) on arbitrary shaped chessboards
A bunch of little data generators to simulate multivariate dependencies.
A small Python script to display the time of your next train(s) in a terminal
my personal website
ELKI Data Mining Toolkit
Code for the paper "Finite-time Analysis of Globally Nonstationary Multi-Armed Bandits", Junpei Komiyama, Edouard Fouché and Junya Honda, 2021.
Solution for the coding challenge Hashcode 2020.
A Scala k-d tree library
(un)official css theme for KIT presentations with jupyter notebook slideshow
Monte Carlo Dependency Estimation
Experiments for the paper "Monte Carlo Dependency Estimation". Fouché & Böhm. SSDBM'19
Experiments for the paper "A framework for dependency estimation in heterogeneous data streams". Fouché et al. - DAPD 2020
Code/Data for the paper "Mining Text Outliers in Document Directories", Fouché et al., ICDM 2020.
Models built with TensorFlow
Experiments about the use of neural networks to discover outliers in high-dimensional data
Simple analysis of the NYC taxi data
Outlier Detection for Text Data
Slides (with sources) for defending the doctoral dissertation "Estimating Dependency, Monitoring and Knowledge Discovery in High-Dimensional Data Streams"
Sources of the doctoral dissertation "Estimating Dependency, Monitoring and Knowledge Discovery in High-Dimensional Data Streams"
A statistical muldi-dimensional stream generator for benchmarking stream mining algorithms
Code and data for the paper "Scaling Multi-Armed Bandit Algorithms". Fouché, Komiyama, Böhm. KDD'19
Automatically exported from code.google.com/p/word2vec