http://bura.brunel.ac.uk/handle/2438/8627 2026-04-20T19:43:01Z 2026-04-20T19:43:01Z Wang, J Yu, K Jiang, R http://bura.brunel.ac.uk/handle/2438/32954 2026-03-11T03:01:05Z 2026-01-13T00:00:00Z

Title: Jiangfeng Wang, Keming Yu and Rong Jiang's contribution to the Discussion of ‘Augmented balancing weights as linear regression’ by Bruns-Smith et al Authors: Wang, J; Yu, K; Jiang, R Abstract: This is an interesting and well-executed paper that makes a substantial contribution to the literature on semiparametric causal inference, elegantly bridging the seemingly distinct fields of balancing weights and regression adjustment. ... Description: Discussion Paper Contribution.

2026-01-13T00:00:00Z Xing, L Kaheh, Z http://bura.brunel.ac.uk/handle/2438/32906 2026-03-01T03:00:47Z 2026-02-24T00:00:00Z

Title: Fair Benchmarking in Short‐Term Load Forecasting Authors: Xing, L; Kaheh, Z Abstract: Performance comparisons in short-term load forecasting are often confounded by differences in preprocessing pipelines rather than reflecting intrinsic architectural capability. Variations in feature engineering, scaling, temporal windowing and data partitioning can dominate reported accuracy and obscure the actual behaviour of forecasting models. This study examines preprocessing–architecture interaction by benchmarking random forest, LightGBM, long short-term memory (LSTM), transformer and Temporal Fusion Transformer (TFT) under a shared tabular preprocessing pipeline, ensuring strict control over data handling and evaluation conditions. Under this controlled setting, tree-based models exhibit strong predictive performance, whereas deep sequence models experience substantial degradation when temporal continuity is not explicitly represented. To isolate architectural sensitivity from preprocessing effects, we further conduct a within-architecture analysis by retraining an identical LSTM under a sequence-aware pipeline aligned with its temporal inductive bias. This realignment yields an order-of-magnitude reduction in RMSE, demonstrating that preprocessing design is a first-order determinant of deep sequence model performance. The results establish a transparent and reproducible benchmarking framework and highlight the importance of aligning data representation with model assumptions when interpreting comparative performance in time series forecasting. Description: Data Availability Statement: The data that support the findings of this study are available from the corresponding author upon reasonable request.

2026-02-24T00:00:00Z Yu, K Jiang, R Wang, J http://bura.brunel.ac.uk/handle/2438/32718 2026-03-09T18:53:16Z 2026-02-18T00:00:00Z

Title: A communication-efficient distributed Retire with application to the analysis of multi-site air-quality distributed data Authors: Yu, K; Jiang, R; Wang, J Abstract: A multi-site city air-quality dataset should be considered distributed data as it is generated from multiple geographically dispersed sources, such as air quality sensors or monitoring stations. In various fields, distributed systems are increasingly employed to handle data collected from diverse sources, often resulting in datasets that are heavy-tailed, asymmetric, or heterogeneous. Robust expectile regression combines the computational efficiency of expectile regression with its robustness in handling heavy-tailed response distributions and outliers. This paper extends robust expectile regression to communication-efficient distributed systems and applies it to the analysis of multi-site air-quality datasets. The proposed distributed estimators achieve both computational and communication efficiency while delivering statistical performance comparable to global estimators, as demonstrated through both theoretical analysis and numerical experiments. Description: Data availability: The air-quality data from the Beijing Municipal Environmental Monitoring Center is available from online site: https://archive.ics.uci.edu/dataset/501/beijing+multi+site+air+quality+data .; Supplementary material: Supplementary data are available online at: https://academic.oup.com/jrsssc/advance-article/doi/10.1093/jrsssc/qlag005/8489473#supplementary-data .

2026-02-18T00:00:00Z Zhang, L Tian, M Yu, K Zhou, M http://bura.brunel.ac.uk/handle/2438/32671 2026-03-09T18:22:43Z 2026-02-05T00:00:00Z

Title: A spatiotemporal marginalized zero-inflated Conway–Maxwell–Poisson regression model: application to international population outmigration within Asia Authors: Zhang, L; Tian, M; Yu, K; Zhou, M Abstract: Asia is a principal source of global migration, and its intra-regional movements profoundly reshape the political, economic, and ecological landscapes of Asian nations. To address the spatiotemporal zero-inflated and dispersion present in migration data, as well as the need for interpretable inference on the overall mean, we develop a spatiotemporal marginalized zero-inflated Conway–Maxwell–Poisson (MZICMP) regression model. This model transcends the limitations of conventional zero-inflated approaches by employing a dispersion parameter that accommodates equidispersion, overdispersion, and under dispersion, and by jointly modelling excess zeros and the marginal mean through the inclusion of country-level covariates, smooth temporal effects, and spatial random effects. For parameter estimation, we implement a Bayesian Markov Chain Monte Carlo algorithm that combines Gibbs sampling with Metropolis–Hastings steps. Simulation demonstrates the model's efficacy in capturing both temporal autocorrelation and spatial zero-inflation patterns, and an empirical application to 1990–2020 intra-Asian out-migration reveals: (1) the share of secondary industry and the share of tertiary industry both show significant negative correlations with out-migration flows, whereas battle-related deaths and the total volume of bilateral trade exhibit positive correlations; (2) the average outmigration trend among Asian countries was relatively high during the period 2005–2010, then declined in 2015–2020; the model results indicate a satisfactory capture of this temporal pattern. Description: Data availability: The data were obtained primarily from the United Nations Population Division (https://population.un.org), the World Bank (https://data.worldbank.org.cn/indicator), and the UCDP database (https://www.pcr.uu.se/research/ucdp). All datasets and codes in this study are available from the corresponding author upon reasonable request.; Supplementary material: Supplementary data are available online at: https://academic.oup.com/jrsssa/advance-article/doi/10.1093/jrsssa/qnag009/8462573?login=true&guestAccessKey=#supplementary-data .

2026-02-05T00:00:00Z