Chien-Min Chen

Dr. Chien-Min Chen

orcid

Research Focus & Metrics

Academic Career Length

1Year (2026-Present)

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

Representative Publications

View All Publications
Verified
Journal Article
Open access
PDF file
Chun-Ta WeiChun-Ta Wei
Chun-Ta Wei
MTMing-Da Tsai
Ming-Da Tsai
YCYu-Lung Chang
Yu-Lung Chang
MWMing-Chih Jason Wang
Ming-Chih Jason Wang
The Full Waveform LiDAR system has been developed and used commercially all over the world. It acts to record the complete time of a laser pulse and has a high-resolution sampling interval compared to the traditional multiple-echo LiDAR, which only provides signals within a single target range. This study area mainly collects data from Riegl LMS-Q680i Full Waveform LiDAR and WorldView-2 satellite imagery, which focuses on buildings, vegetation, grassland, asphalt roads and other ground types as the surface objects. The amplitude and pulse width are selected as waveform basic parameters. The parameter of topography is slope, and the height classification parameters of the test ground are 0–0.5 m, 0.5–2.5 m, and 2.5 m. To eliminate noise, the neighborhood average is applied on the LiDAR parameter values and analyzed as the classification accuracy comparison. This survey uses Decision Tree as the classification method. Comparing the data between neighborhood average and non-neighborhood average, the data classification accuracy improves by 7%, and Kappa improves by 5.92%. NDVI image data are utilized to distinguish the artificial from natural ground. The results show that the neighborhood average with previous data can improve the classification accuracy by 5%, and Kappa improves by 4.25%. By adding NIR-2 of WorldView-2 satellite imagery to the neighborhood average analysis, the overall classification accuracy is improved by 2%, and the Kappa value by 1.21%. This article shows that utilizing the analysis of neighborhood average and image parameters can effectively improve the classification accuracy of land covers.
1

23 September 2026 in Sustainability
Verified
Journal Article
Open access
PDF file
MWMing-Chih Jason Wang
Ming-Chih Jason Wang
CCChien-Min Chen
Chien-Min Chen
Urban inundation forecasting in high-density cities is complicated by sparse incident labels, nonstationary rainfall, and changing sensing systems. Using Taipei as a case study, we develop a building-scale GeoAI decision-support prototype that integrates 30,421 Level of Detail 1 (LOD1) building-centroid nodes, 41,636 graph edges, built-environment attributes, and eight extreme-rainfall events from 2015 to 2026. We evaluate a static inundation-susceptibility baseline, 60 and 120 min event-level forecasts, tree-ensemble models, a Temporal GNN, and ConvLSTM and separately test temporal and spatial transfer under compound distribution shift. The analysis treats the data as having positive–unlabeled characteristics and distinguishes event-level, pooled, temporal-transfer, spatial-transfer, and scale-sensitivity evaluation regimes. All stochastic models were retrained under five fixed random seeds. Across the six primary events (the simulation-dominant E2016_Megi event excluded as a sensitivity case), the five-seed macro-averaged AUCs were 0.651 ± 0.035 at 60 min (95% event-cluster bootstrap CI 0.533–0.789) and 0.626 ± 0.016 at 120 min (0.445–0.805); including E2016_Megi raises the seven-event means to 0.694 ± 0.030 and 0.672 ± 0.014. Under non-equivalent spatial supports, the five-seed benchmark yields random forest AUC 0.711 ± 0.005, XGBoost 0.695 ± 0.000, Temporal LSTM 0.622 ± 0.009, Temporal GNN 0.477 ± 0.011, and ConvLSTM 0.408 ± 0.044; the ranking is configuration-specific and does not establish architecture superiority, whereas forward temporal validation falls to 0.549 ± 0.062. Physics-based synthetic augmentation increased the mean AUC for four historical gauge-era events only modestly, from 0.651 ± 0.013 to 0.668 ± 0.009, and one event deteriorated. These results define the conditions under which building-scale GeoAI remains informative and where it fails, providing a more defensible basis for urban-resilience decision support, climate adaptation, and infrastructure continuity.

28 August 2026 in Preprints.org
Verified
Preprint
Open access
PDF file
MWMing-Chih Jason Wang
Ming-Chih Jason Wang
CCChien-Min Chen
Chien-Min Chen
Urban inundation forecasting in high-density cities is complicated by sparse incident labels, nonstationary rainfall, and changing sensing systems. Using Taipei as a case study, we develop a building-scale GeoAI decision-support prototype that integrates 30,421 LOD1 building-centroid nodes, 41,636 graph edges, built-environment attributes, and eight extreme-rainfall events from 2015 to 2026. We evaluate a static inundation-susceptibility baseline, 60- and 120-min event-level forecasts, tree-ensemble models, a Temporal GNN, and ConvLSTM, and separately test temporal and spatial transfer under compound distribution shift. The analysis adopts a positive-unlabeled (PU) learning perspective and distinguishes event-level, pooled, temporal-transfer, spatial-transfer, and scale-sensitivity evaluation regimes. Across seven events with verifiable outputs, macro-averaged AUCs were 0.694 at 60 min and 0.677 at 120 min. XGBoost achieved the highest AUC (0.682) in a benchmark with non-equivalent spatial support, whereas forward temporal validation fell to 0.490. Physics-based synthetic augmentation increased the mean AUC for four historical gauge-era events from 0.4965 to 0.6768, although one event deteriorated. These results define the conditions under which building-scale GeoAI remains informative and where it fails, providing a more defensible basis for urban-resilience decision support, climate adaptation, and infrastructure continuity.

14 August 2026 in Preprints.org
Verified
Preprint
Open access
PDF file
MWMing-Chih Jason Wang
Ming-Chih Jason Wang
CCChien-Min Chen
Chien-Min Chen
Compact development is often treated as a low-carbon strategy, yet high-density cities may face new carbon pressures as vertical development, agglomeration, and cross-district commuting intensify. This study examines Taipei using 27,239 100 m × 100 m grid cells and 1980–2025 data integrating Historical GIS, remote sensing, official LOD1 building models, energy statistics, and anonymized commuting origin–destination data. XGBoost, TreeSHAP, multiscale geographically weighted regression (MGWR), CASA, and network analysis served as complementary instruments. The carbon-emission compound annual growth rates of existing built-up cells were 2.994% in the historic western districts and 2.890% in the emerging eastern districts (p = 0.104); the larger long-term increase in the east (approximately 195% versus 166%) is more consistent with extensive-margin expansion. TreeSHAP identified breakpoints at 12.83 for nighttime lights, 197.67 buildings/km² for building density, and 99.61 for FAR × Nightlight. MGWR achieved an adjusted R² of 0.886 and revealed marked multiscale spatial non-stationarity. Commuting flows further indicated cross-boundary carbon-burden asymmetry. The findings recast compact-city benefits as contingent on development margins, morphological thresholds, spatial scale, and functional urban networks. As an observational reconstruction, the study supports predictive and spatial associations rather than strict causal effects.

22 September 2026 in Sustainability
Verified
Journal Article
Open access
PDF file
MWMing-Chih Jason Wang
Ming-Chih Jason Wang
MWMing-Chih Jason Wang
Ming-Chih Jason Wang
Compact development is often treated as a low-carbon strategy; yet, high-density cities may face new carbon pressures as vertical development, agglomeration, and cross-district mobility intensify. This study examines Taipei using 27,239 100 m × 100 m grid cells and 1980–2025 data integrating Historical GIS, remote sensing, official LOD1 building models, zone-level energy statistics, and anonymized origin–destination data. Zone-level energy totals were converted to CO2 and allocated to 100 m cells through a mass-preserving proxy surface; the resulting grid target is therefore a reconstructed allocation rather than an independently observed cell-level outcome. XGBoost, TreeSHAP, multiscale geographically weighted regression (MGWR), CASA, and network analysis served as complementary instruments. Random ten-fold cross-validation provided internal reconstruction diagnostics but was not spatially blocked and does not establish spatial transferability. The mean carbon-emission compound annual growth rates of existing built-up cells were 2.994% in the historic western districts and 2.890% in the emerging eastern districts (Welch’s test, p = 0.104). This non-significant result does not establish equivalence or identify the contributions of existing and newly built-up areas to aggregate growth; the extensive-margin hypothesis remains untested. TreeSHAP identified model-specific empirical breakpoints at 12.83 for a standardized, unitless nighttime-light index, 197.67 buildings/km2 for the building density, and 99.61 for the dimensionless FAR × Nightlight composite. These values indicate changes in the model-predicted feature contributions; they are not physical emission triggers, causal transition points, universal planning standards, or transferable regulatory cut-offs. MGWR achieved an adjusted R2 of 0.886 and revealed marked multiscale spatial non-stationarity; its coefficients are interpreted only as conditional spatial statistical associations. The observed Taipei Metro OD flows further indicated a cross-boundary carbon-burden asymmetry, but the network covers metro travel only and excludes buses, motorcycles, private cars, walking, cycling, and other modes. As an observational and proxy-based reconstruction, the study supports predictive and spatial associations rather than strict causal effects.