A5036115322- Tensor decomposition and applications
- Matrix Theory and Algorithms
- Ecology and Vegetation Dynamics Studies
- Species Distribution and Climate Change
- Land Use and Ecosystem Services
- Remote Sensing in Agriculture
- Statistical and numerical algorithms
- Data Analysis with R
- Plant and animal studies
- Theoretical and Computational Physics
- Blind Source Separation Techniques
- Geochemistry and Geologic Mapping
- Computational Physics and Python Applications
- Sensory Analysis and Statistical Methods
- Control Systems and Identification
- Distributed and Parallel Computing Systems
- Economic and Technological Innovation
- Neural Networks and Applications
- Optical measurement and interference techniques
- Leaf Properties and Growth Measurement
- Advanced Neuroimaging Techniques and Applications
- Electromagnetic Scattering and Analysis
- Advanced Optimization Algorithms Research
- Advanced Mathematical Theories and Applications
- Medical Image Segmentation Techniques
Centre de Recherche Inria Bordeaux - Sud-Ouest
2019-2022
Numerical Algorithms Group (United Kingdom)
2021
Institut national de recherche en informatique et en automatique
2019
Ecosystem heterogeneity has been widely recognized as a key ecological indicator of several functions, diversity patterns and change, metapopulation dynamics, population connectivity or gene flow.In this paper, we present new R package-rasterdiv-to calculate indices based on remotely sensed data. We also provide an application at the landscape scale demonstrate its power in revealing potentially hidden patterns.The rasterdiv package allows calculating multiple indices, robustly rooted...
Abstract Aim The majority of work done to gather information on the Earth's biodiversity has been carried out using in‐situ data, with known issues related epistemology (e.g., species determination and taxonomy), spatial uncertainty, logistics (time costs), among others. An alternative way about ecosystem variability is use satellite remote sensing. It works as a powerful tool for attaining rapid standardized information. Several metrics used calculate remotely sensed diversity ecosystems...
The efficient solution of large-scale multiterm linear matrix equations is a challenging task in numerical algebra, and it largely open problem. We propose new iterative scheme for symmetric positive definite operators, significantly advancing methods such as truncated matrix-oriented Conjugate Gradients (CG). algorithm capitalizes on the low-rank format its iterates by fully exploiting subspace information factors iterations proceed. approach implicitly relies orthogonality conditions...
Abstract The variation of species diversity over space and time has been widely recognised as a key challenge in ecology. However, measuring large areas might be difficult for logistic reasons related to both cost savings sampling, well accessibility remote ecosystems. In this paper, we present new package - calculate indices based on remotely sensed data, by discussing the theory behind developed algorithms. Obviously, measures from should not viewed replacement situ data biological...
Abstract We propose a new method to estimate plant diversity with Rényi and Rao indexes through the so called High Order Singular Value Decomposition (HOSVD) of tensors. Starting from NASA multi-spectral images we evaluate compare original estimates those realized via HOSVD compression methods for big data. Our strategy turns out be extremely powerful in terms memory storage precision outcome. The obtained results are promising that can support efficiency our ecological framework.
Abstract Aim The majority of work done to gather information on Earth diversity has been carried out by in-situ data, with known issues related epistemology (e.g., species determination and taxonomy), spatial uncertainty, logistics (time costs), among others. An alternative way about ecosystem variability is the use satellite remote sensing. It works as a powerful tool for attaining rapid standardized information. Several metrics used calculate remotely sensed ecosystems are based Shannon’s...
Abstract The variation of species diversity over space and time has been widely recognised as a key challenge in ecology. However, measuring large areas might be difficult for logistic reasons related to both cost savings sampling, well accessibility remote ecosystems. In this paper, we present new R package - rasterdiv calculate indices based on remotely sensed data, by discussing the theory beyond developed algorithms. Obviously, measures from should not viewed replacement in-situ data...
We consider the solution of linear systems with tensor product structure using a GMRES algorithm. In order to cope computational complexity in large dimension both terms floating point operations and memory requirement, our algorithm is based on low-rank representation, namely Tensor Train format. backward error analysis framework, we show how approximation affects accuracy computed solution. With bacwkward perspective, investigate situations where $(d+1)$-dimensional problem be solved...
In the framework of tensor spaces, we consider orthogonalization kernels to generate an orthogonal basis a subspace from set linearly independent tensors. particular, experimentally study loss orthogonality six methods, namely Classical and Modified Gram-Schmidt with (CGS2, MGS2) without (CGS, MGS) re-orthogonalization, Gram approach, Householder transformation. To overcome curse dimensionality, represent tensors low-rank approximation using Tensor Train (TT) formalism. addition, introduce...
Abstract Ecosystem heterogeneity has been widely recognized as a key ecological feature, influencing several functions, since it is strictly related to functions like diversity patterns and change, metapopulation dynamics, population connectivity, or gene flow. In this paper, we present new R package - rasterdiv calculate indices based on remotely sensed data. We also provide an application at the landscape scale demonstrate its power in revealing potentially hidden patterns. The allows...