Guilherme Coelho

Mr. Guilherme Coelho


Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
23 October 2026 in SSRN Electronic Journal
Verified
Preprint
Open access
GCGuilherme Coelho
Guilherme Coelho
CCClarimar José Coelho
Clarimar José Coelho
Learned per-wavelength weights are widely used to make chemometric neural networks interpretable. A common design passeseach weight through a sigmoid and initialises it from domain knowledge, expecting training to refine the prior against the spectra.We ask which weights the data can inform. The gradient reaching a weight is attenuated by s(1- s), so weights initialised near eitherend of the unit interval are effectively insulated from the data over realistic training budgets; where the gradient does arrive, spectralcollinearity constrains combinations of neighbouring weights rather than individual ones. We analyse a prior-guided spectral gatingmodel on near-infrared and Raman data, with synthetic verification against known ground truth. Over the bands given the maximumprior weight the data supply no measurable curvature — a contraction ratio of unity across the approximations, reference scales,optimisers and fitted solutions examined — while the curvature eigenvectors show the data determining regional combinationsof weights and leaving trades between neighbours free. Neither bound appears in predictive performance: with realistic spectralcorrelation, a prior contradicting the true relevance matches the correct one to within a pre-specified margin of 0.02 in held-out R2while placing 2% of the gate weight on the relevant regions against 72%; region-level permutation shows the network relying onthose regions regardless. Under the training conditions examined, prior specification therefore bounds what spectral attention canlearn, and predictive accuracy cannot validate the explanation. Normalising a prior by its maximum turns a flat top into a block ofinert parameters. Keeping the prior’s weights away from the extremes raises the agreement between the gate and an operationalreference for band relevance — the absolute coefficients of a calibrated partial least squares model, not a chemical ground truth —in every one of five partitions grouped by harvest population; the accompanying advantage in held-out accuracy does not survivethe same grouping, and we report it as specific to the partition on which it was measured.

7 February 2026 in Climate
Verified
Journal Article
Open access
PDF file
DKDanielle Satie Kassada
Danielle Satie Kassada
LSLarissa Marques Suardi
Larissa Marques Suardi
IRIgor de Lima Peixoto Rocha
Igor de Lima Peixoto Rocha
GCGuilherme Coelho
Guilherme Coelho
APAna Carolina Souza Peratelli
Ana Carolina Souza Peratelli
Ambient air pollution is a major environmental health risk, with impacts that may be unevenly distributed across vulnerable populations, particularly in urban areas under changing climate conditions. This study assessed whether social conditions modify the association between exposure to fine particulate matter (PM2.5) and deaths from circulatory and respiratory diseases in Campinas, Brazil. An ecological time-series study was conducted from 2010 to 2023 using monthly death data aggregated at the catchment-area level of 63 Primary Health Care Centers, stratified into four vulnerability groups. Associations between monthly PM2.5 concentrations and mortality rates were estimated using distributed lag non-linear models within a generalized additive modeling framework, adjusting for meteorological variables, ozone, seasonality, and long-term trends. A total of 38,311 cardiorespiratory deaths were recorded. Mortality rates varied non-linearly across strata, and a statistically significant positive association between PM2.5 exposure and mortality was observed only in the intermediate–high vulnerability group at higher pollution levels. No significant associations were identified in the lowest or highest vulnerability strata. These findings indicate that social context modifies air pollution–related mortality risks, highlighting the importance of incorporating vulnerability dimensions into air quality assessments and climate-related public health strategies.

11 October 2025 in COVID
Verified
Journal Article
Open access
PDF file
DKDanielle Satie Kassada
Danielle Satie Kassada
IRIgor de Lima Peixoto Rocha
Igor de Lima Peixoto Rocha
GCGuilherme Coelho
Guilherme Coelho
APAna Carolina Souza Peratelli
Ana Carolina Souza Peratelli
Severe acute respiratory syndrome (SARS), caused by the COVID-19 virus, continues to pose a significant public health challenge in Brazil, particularly in 2024, with high mortality rates among vulnerable groups. This study aimed to describe the sociodemographic, clinical and vaccination profiles of SARS cases due to COVID-19 in Brazil in 2024, as well as analysing factors associated with clinical outcomes such as death, admission to the intensive care unit (ICU) and the need for ventilatory support. A total of 30,529 reported cases were analysed. On average, the interval between the last vaccine dose and symptom onset was 30.31 months (SD = 6.77), while the interval between symptom onset and clinical outcome was 13.26 days (SD = 16.55), revealing significant variability. The results showed higher mortality rates among men (23.7%) than women (19.1%) (p < 0.0001). Mortality increased progressively with age, reaching 24.4% in individuals aged 60 years or older, whereas rates were below 2% in children under 10 years of age. The highest proportions of deaths were observed regionally in the Northeast (26.8%) and North (22.6%), in contrast to the Midwest (17.7%) (p < 0.0001). Men were also more likely to require ICU admission (38.1% vs. 33.6%) and ventilation (62.9% vs. 60.5%). A time interval of over 24 months since the last vaccine dose was associated with higher mortality (21.9% vs. 20.6%; p = 0.0005). These results highlight the importance of ongoing surveillance and updating the vaccination schedule, particularly for more vulnerable populations.

21 December 2025 in EarthArXiv
Verified
Preprint
Open access
PDF file
GCGuilherme Coelho
Guilherme Coelho
CCCharles M'POca Charles
Charles M'POca Charles
CCClarimar José Coelho
Clarimar José Coelho
MJMauricio Wesley Perroud Junior
Mauricio Wesley Perroud Junior
DKDanielle Satie Kassada
Danielle Satie Kassada
RPRodolfo de Carvalho Pacagnella
Rodolfo de Carvalho Pacagnella
We conducted a nationwide ecological time-series study to quantify the association between ambient temperature and respiratory mortality across Brazil's diverse climates. Using data from 520 municipalities (population ≥50,000) from 2010 to 2020, we analysed 1,008,157 respiratory deaths (ICD-10 J00–J99) using distributed-lag non-linear models (lags 0–21 days) with quasi-Poisson regression, adjusting for seasonality, time trends, and day of week. City-specific estimates were pooled by random-effects meta-analysis, and attributable fractions were calculated using the forward perspective method. Both cold and heat were associated with increased mortality risk, following a U-shaped relationship. Minimum mortality temperature varied by 6.3°C across regions. Relative risks at the 10th and 90th percentiles were 1.95 (IQR 1.44–2.88) and 1.82 (IQR 1.39–2.81), respectively. Non-optimal temperatures accounted for 2.89% of deaths nationally (approximately 2,650 annually), with the highest fractions in the North (8.2%). Adults aged ≥65 years and Black individuals (for heat) showed the most significant vulnerability. Temperature extremes increase respiratory mortality across Brazil, with tropical regions exhibiting heightened vulnerability to cold despite warmer baseline climates. These findings support region-specific warning systems addressing both thermal extremes, with priority for older adults and Black populations.

7 July 2026 in SSRN Electronic Journal
Verified
Preprint
Open access
GCGuilherme Coelho
Guilherme Coelho
LCLilian Carla Carneiro
Lilian Carla Carneiro
AFArlindo Rodrigues Galvão Filho
Arlindo Rodrigues Galvão Filho
CCClarimar José Coelho
Clarimar José Coelho
Prior-Guided Spectral Gating (PGSG) incorporates domain knowledge intonear-infrared (NIR) regression through a learnable gate vector initializedfrom a spectroscopic prior. We identify two fundamental limitations of theoriginal formulation that prevent its application to high-dimensional spectra:a softmax gradient vanishing problem (at p = 281 bands, the softmax Jaco-bian diagonal is 58× smaller than sigmoid, rendering the gate untrainable),and a circular prior pitfall (priors derived from training data cause the gateto reproduce the prior without adaptation). We propose PGSGv2, which re-places softmax with independent per-band sigmoid gates and a PLS regres-sor with an end-to-end MLP, enabling backpropagation through the jointgate–regressor objective. Applied to mango dry matter content (DMC%)prediction using the Mango DMC v3 dataset (n = 1,448, p = 281 effec-tive bands, within-season protocol) with a literature-based prior, PGSGv2achieves R2 = 0.806 at n = 1,159, outperforming PLS (R2 = 0.568, +42%)and MLP (R2 = 0.699, +15%). For n ≤ 60, PGSGv2 is the only methodwith positive R2 , demonstrating the value of prior-guided initialization indata-scarce regimes. Learned gates are chemically coherent (ρ(g, s) ≥ 0.998)and highly reproducible across seeds (Jaccard ≥ 0.945). PGSGv2 providesa scalable and interpretable architecture for NIR calibration when labeledsamples are limited.

7 August 2026 in SSRN Electronic Journal
Verified
Preprint
Open access
GCGuilherme Coelho
Guilherme Coelho
CCClarimar José Coelho
Clarimar José Coelho
Prior-Guided Spectral Gating (PGSG) is a chemometrically-informed spec-tral attention mechanism, previously validated only on near-infrared (NIR)spectroscopy, where it outperforms partial least squares (PLS) at scale (n ≈1,448, R2 = 0.806 versus 0.568). Whether the mechanism generalizes tomodalities with markedly different spectral structure, such as Raman, hasnot been tested. We evaluate the unmodified architecture — an independentsigmoid gate per variable feeding a shallow multilayer perceptron, trainedend-to-end — on public NIR (n = 1,448, dry-matter content) and Raman(n = 6,472, glucose concentration) benchmarks, and formalize three descrip-tors of learned-gate topology (entropy, smoothness, sparsity) to characterizecross-modality structural differences. The model outperforms PLS in bothmodalities (∆R2 PLS = +0.244 NIR, +0.025 Raman); Raman gates are moresparse than NIR gates, and NIR gates are smoother than Raman gates oncesmoothness is normalized by physical band density, a correction we show tobe necessary and that reverses the naive raw-index comparison. A literature-informed prior does not improve accuracy over an uninformed initializationbut substantially increases gate reproducibility across training seeds (pairwiseJaccard similarity 0.90–0.94 versus 0.58–0.65). A finer, band-level test corre-lating local gate structure with literature-reported vibrational linewidths did not yield a robust correlation. These results indicate that PGSG generalizesacross spectroscopic modalities without architectural modification, that themagnitude of its advantage scales with how diluted the analyte signal alreadyis in a given modality, and that the principal value of a physically motivatedprior is interpretive stability rather than predictive gain.