ULCO-LISIC – classification and modeling tools for data analysis and prediction.
Short title2: LISIC clusTools
Link to website: https://mawenzi.univ-littoral.fr/RclusTool/
Mawenzi is a tool center with R-packages and their RShiny graphical user interfaces for data interpretation, from data completion to data prediction.
RClusTool: Clustering and Classification Tool, with Visualisation and Labelling Features. R-package and its GUI are well adapted to clustering, insert expert knowledge in this clustering as label or pair constraints (these objects must be linked or not linked). Flow cytometry data with pulse signals and/or features could be used with an image visualisation to help in the expert classification step. But it is generalised to any dataset with at least feature input. This package is developed by LISIC-ULCO (Pierre-Alexandre Hébert, Emilie Poisson) in collaboration with CNRS-LOG (Luis Felipe Artigas).
uHMM: Unsupervised Hidden Markov Model, automatic segmentation of time series. This R-package and its GUI are well adapted to segment times series in recurrent and rare/extreme events and to model their dynamics. It offers many data exploration, visualisation reports for data managers and stakeholders. Marine time series from fixed or mobile buoys could be used. A MAREL Carnot data sample is provided in this tool for application.
sClust: Spectral Clustering toolbox and GUI allow to detect populations or events with a scalable point of view from several state-of-art clustering algorithms with spectral clustering focus.
Marine time series or dataframe could be used.
DTWBI: Univariate signal – Dynamic Time Warping based Imputation, filling large gaps within time series. DTWBI is a R package to complete time series where a cause-effect relationship exists within the dynamics process. It is well adapted to large gap (no-consecutive values) and time series with low auto-correlation.
Université du Littoral Côte d’Opale, Laboratoire d’Informatique Signal et Image de la Côte d’Opale
Contact name: Emilie Poisson Caillault
Location: Université du Littoral Côte d’Opale - Laboratoire d’Informatique Signal et Image de la Côte d’Opale - UR 4491, Maison de la Recherche Blaise Pascal, 50 rue Ferdinand Buisson, BP 719 62228 Calais Cedex, France
Clustering, Classification, Constraint Classification, Time series completion, Hidden Markov Modeling, Visualisation tool, Interpretation tool, Data analysis