Xuankai Zhang

Dr. Xuankai Zhang

orcid

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
12 November 2025 in Systems
Verified
Journal Article
Open access
PDF file
Pei DuPei Du
Pei Du
XZXuankai Zhang
Xuankai Zhang
TCTingting Chen
Tingting Chen
Wendong YangWendong Yang
Wendong Yang
Scientific, accurate, and interpretable carbon price forecasts provide critical support for addressing climate change, achieving low-carbon goals, and informing policy-making and corporate decision-making in energy and environmental markets. However, the existing studies mainly focus on deterministic forecasting, with obvious limitations in data feature diversity, model interpretability, and uncertainty quantification. To fill these gaps, this study constructs an interpretable hybrid system for carbon market price prediction by combining feature screening algorithms, deep learning models, and interpretable explanatory analysis methods. Specifically, this study first screens important variables from twenty-one multi-source structured and unstructured influencing factor datasets on five dimensions affecting carbon price using the Boruta algorithm. Immediately after that, this study proposes a hybrid architecture of bidirectional temporal convolutional network and Informer models, where a bidirectional temporal convolutional network is used to extract local spatio-temporal dependent features, while Informer captures long sequences of global features through the connectivity mechanism, thus realizing staged feature extraction. Then, to improve the interpretability of the model and quantify the uncertainty, this study introduces Shapley additive explanations to analyze the feature contribution in the prediction process, and the Monte Carlo dropout method is used to achieve interval prediction. Finally, the empirical results in China’s Guangdong and Shanghai carbon markets show that the proposed model significantly outperforms benchmark models, and the coverage probability of the obtained prediction intervals significantly outperforms the confidence level. The Shapley additive explanation analysis reveals regional heterogeneity drivers. In addition, the proposed model is also intensively validated in the European carbon market and the U.S. natural gas market, which also demonstrate an excellent prediction performance, indicating that the model has good robustness and applicability.

20 October 2026 in SSRN Electronic Journal
Verified
Preprint
Open access
JHJinghan Hu
Jinghan Hu
JZJiarui Zhang
Jiarui Zhang
CLChengxi Li
Chengxi Li
CLChengkun Luo
Chengkun Luo
XZXuankai Zhang
Xuankai Zhang
WYWeiyao Yang
Weiyao Yang
Metro automatic fare collection (AFC) is data-rich but state-incomplete. Fare gates record entries and exits, but not line-level passenger states or in-station transfers. We cast passenger-flow reconstruction as reliability-aware state estimation under incomplete fare sensing and follow the estimate through uncertainty validation into selective decision-making. The evaluation draws on 123 days of publicly available Shanghai Metro records, over 267 million trips across 302 stations and 14 lines. First, the directly constrained component is reconstructed in closed form from stable empirical alighting shares, attaining a pooled direct alighting weighted absolute percentage error (WAPE) of 0.055 against smartcard ground truth, a level that learned representations do not reach at any tested capacity. Second, a single shared between-day dispersion factor substantially improves empirical coverage of nominal 95% intervals, from 0.073–0.229 to 0.515–0.874 on the four evaluated lines, pointing to day-to-day volatility, not fitting error, as the leading contributor to posterior overconfidence. Third, posterior reliability is rank-consistent but not level-calibrated. Across 114 line–interchange interfaces it orders true reconstruction risk under a transfer mechanism the model did not assume (rank correlation 0.495, permutation 𝑝 = 0.0005), and the ordering survives the removal of real observations. Fourth, the ranking supports selective decisions. Rejecting the least reliable states and steering additional sensing by posterior uncertainty captures 62–74% of the oracle error reduction at small budgets. The framework thus reframes origin–destination (OD) estimation as partially observable passenger-state reconstruction in which every estimate carries a reliability that guides selective inference and sensing