A Simulation Study to Evaluate Bayesian LASSO’s Performance in Zero-Inflated Poisson (ZIP) Models
Bibliographic record
Abstract
When modelling count data, it is possible to have excessive zeros in the data in many applications. My thesis concentrates on the variable selection in zero-inflated Poisson (ZIP) models. This thesis work is motivated by Brown et al. (2015), who considered the excessive amount of zero in their data structure and the site-specific random effects, and used Bayesian LASSO method for variable selection in their post-fire tree recruitment study in interior Alaska, USA and north Yukon, Canada. However, the above study has not carried out systematic simulation studies to evaluate Bayesian LASSO’s performance under different scenarios. Therefore, my thesis conducts a series of simulation studies to evaluate Bayesian LASSO’s performance with respect to different setting of some simulation factors. My thesis considers three simulation factors: the number of subjects (N), the number of repeated measurements (R) and the true values of regression coefficients in the ZIP models. With different settings of the three factors, the proposed Bayesian LASSO’s performance would be evaluated using three indicators: the sensitivity, the specificity and the exact fit rate. For applied practitioners, my thesis would be a useful example demonstrating under what circumstances one can expect Bayesian LASSO to have good performance in ZIP models. After sorting out the simulation results, we can find that Bayesian LASSO’s performance is jointly affected by all the three simulation factors, while this method of variable selection is more reliable when the true coefficients are not close to zero. My thesis also has some limitations. Primarily, with the time limitation of my thesis, it is impossible to consider all the factors that can potentially affect the simulation results, and using other penalty forms other than L1 penalty is also left for future researchers to work on. Moreover, the current variable selection method is only for fixed effects selection while the variable selection for the mixed effect selection in ZIP models can be a direction for future work.
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.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.021 | 0.046 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.002 | 0.003 |
| Insufficient payload (model declined to judge) | 0.002 | 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 itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".