PostDoc Project
Machine learning mathods for spatio-temporal data fusion
2021/12/01 - 2022/11/30
The general goal of the project is relating to the development of machine learning methods to represent, analyze and process spatio-temporal data. The
targeted application is cell biology; more precisely the spatio-temporal analysis of biological mechanism in cell populations. The developments will focus
on i) the proposal of original multimodal data fusion algorithms, ii) their performance analysis and iii) their application to real data sets. The proposed
approaches will make use of advanced techniques related to tensor decompositions and/or deep neural networks. A special attention will be paid to the
design of methods robust to inter and intra individual variability, allowing to handle large size datasets. All the application part will be made in close
collaboration with biologists from the CRAN.

The candidate should demonstrate methodological and/or theoretical skills in one or more of the following areas:
- Theory of deep neural networks
- Low rank and/or parsimonious models
- Signal processing on graphs
machine learning, data fusion,
Biology, Signals and Systems in Cancer and Neuroscience