Exploiter l'approche hiérarchique bayésienne pour la modélisation statistique de structures spatiales. Application en écologie des populations
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
For most ecological questions, the random processes studied are spatially structured and come from the combined effect of several observed or unobserved random variables interacting at various scales. In practice, when data can't be directly treated with traditional spatial structures, observations are often considered as independent. Moreover, the usual models are often based on hypotheses that are too simple with regards to the complexity of the studied phenomena. In the present work, the hierarchical modelling framework is combined with some spatial statistics tools to build specific functional random structures for complex and spatially structured phenomena in population ecology. Model inference is done under the bayesian framework using MCMC algorithms. In the first part, a spatial hierarchical model (called Geneclust) is developed to identify genetically homogeneous populations when genetic diversity varies continuously in space. A hidden Markov random field, used to model the spatial structure of genetic diversity, is combined with a bivariate model for the occurrence of genotypes to take into account the possible occurrence of inbreeding in some natural populations. In the second part of the thesis, a particular compound Poisson process, called law of leaks, is presented from the hierarchical point of view. The goal was to describe the process of sampling living organisms. This approach explicitly confronts the technical issue of modelling continuous zero-inflated data from sampling characterized many zero values and variable sampling effort. This model is combined with different area-based models to add spatial dependencies between geographical units then with a bivariate gaussian random field built by process convolutions to model the joint spatial distribution of two species. The fitting and predictive capacities of the different hierarchical models are compared to the traditional models from simulated and real data (Scandinavian brown bears, epibenthic invertebrates in Saint-Lawrence Gulf (Canada)
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it