Yixuan Y. Zheng
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Yixuan Y. Zheng has a total of 12 co-authors. Here are the top co-authors by number of shared publications:
- Misako Takayasu
Institute of Science Tokyo
3 shared publications Yohei Shida
School of Computing, Institute of Science Tokyo
2 shared publications- Arthur Matsuo Yamashita Rios de Sousa
Department of Mathematical and Computing Science, School of Computing, Tokyo Institute of Technology, 4259, Nagatsuta-cho, Yokohama 226-8502, Japan
1 shared publication Daisuke Ando
School of Computing, Institute of Science Tokyo
1 shared publicationJiwei J. Jiang
School of Computing, Institute of Science Tokyo
1 shared publicationJunichi Ozaki
Yokohama City University
1 shared publication
Latest Publications
- Open Access
- 3 November 2026
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 21 (11) , e0344222
Weekly sales forecasting using time series decomposition with irregular seasonality: A case study of Sony TV sales in the United States
In the practice of sales forecasting, alongside high accuracy, simplicity and interpretability are desirable qualities in a predictive model. Here, we use time series decomposition to construct a simple and interpretable forecasting method for weekly sales, taking as a case study the total weekly sales of Sony TV units in the United States. The main challenge is to properly address irregular seasonality arising from the way a year is divided into weeks and from floating holidays and events. As a solution, we propose the concept of ‘analogous weeks’, a general framework for grouping weeks that are similar according to some criteria, which encodes seasonality and already accounts for calendar effects. By incorporating this notion into the time series decomposition and forecasting, we achieved yearly average percentage errors of less than 10% for the 1-week-ahead forecast and less than 15% for forecast horizons of up to 26 weeks, accuracy levels comparable to or better than those of other established seasonal models, namely Prophet and SARIMAX, particularly at longer forecast horizons.
- 3 November 2026
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 21 (11) , e0344222
Open AccessWeekly sales forecasting using time series decomposition with irregular seasonality: A case study of Sony TV sales in the United States
In the practice of sales forecasting, alongside high accuracy, simplicity and interpretability are desirable qualities in a predictive model. Here, we use time series decomposition to construct a simple and interpretable forecasting method for weekly sales, taking as a case study the total weekly sales of Sony TV units in the United States. The main challenge is to properly address irregular seasonality arising from the way a year is divided into weeks and from floating holidays and events. As a solution, we propose the concept of ‘analogous weeks’, a general framework for grouping weeks that are similar according to some criteria, which encodes seasonality and already accounts for calendar effects. By incorporating this notion into the time series decomposition and forecasting, we achieved yearly average percentage errors of less than 10% for the 1-week-ahead forecast and less than 15% for forecast horizons of up to 26 weeks, accuracy levels comparable to or better than those of other established seasonal models, namely Prophet and SARIMAX, particularly at longer forecast horizons.
- Open Access
- 24 June 2026
- Published by Scilight Press Pty Ltd in Journal of Social Physics
- Vol. 1 (1) , 6
Do Railway Commuters Exhibit Consistent Route Choice “Rationality” Across Different Contexts and Time? Evidence from Tokyo Metropolitan Commutes
In urban railway systems, if every commuter were perfectly rational, concentrated demand on optimal routes would undermine system performance. Understanding the actual degree of collective route choice determinism, and whether it remains stable over time, is essential for transportation policy. Smartphone GPS data offers both the scale and complete trajectory coverage that survey data and smart card data individually lack, yet its coarse spatial accuracy has limited its adoption for railway route identification. To overcome this, we develop a methodology to identify railway commuting routes from one year of GPS data covering over one million daily users in the Tokyo metropolitan area. We apply a multinomial logit (MNL) framework, utilizing a standardization convention to extract a comparative measure of collective determinism alongside relative attribute preferences. We find that collective route choices in Tokyo are strongly cost-sensitive but non-deterministic. Both components remain stable across all twelve months of 2023, confirming that this determinism represents a structural consistency of the commuting system. Leveraging the dataset’s scale, we further estimate parameters for each origin–destination (OD) pair individually, revealing systematic heterogeneity driven by departure time and transport complexity that smaller datasets treat as unobserved.
- 24 June 2026
- Published by Scilight Press Pty Ltd in Journal of Social Physics
- Vol. 1 (1) , 6
Open AccessDo Railway Commuters Exhibit Consistent Route Choice “Rationality” Across Different Contexts and Time? Evidence from Tokyo Metropolitan Commutes
In urban railway systems, if every commuter were perfectly rational, concentrated demand on optimal routes would undermine system performance. Understanding the actual degree of collective route choice determinism, and whether it remains stable over time, is essential for transportation policy. Smartphone GPS data offers both the scale and complete trajectory coverage that survey data and smart card data individually lack, yet its coarse spatial accuracy has limited its adoption for railway route identification. To overcome this, we develop a methodology to identify railway commuting routes from one year of GPS data covering over one million daily users in the Tokyo metropolitan area. We apply a multinomial logit (MNL) framework, utilizing a standardization convention to extract a comparative measure of collective determinism alongside relative attribute preferences. We find that collective route choices in Tokyo are strongly cost-sensitive but non-deterministic. Both components remain stable across all twelve months of 2023, confirming that this determinism represents a structural consistency of the commuting system. Leveraging the dataset’s scale, we further estimate parameters for each origin–destination (OD) pair individually, revealing systematic heterogeneity driven by departure time and transport complexity that smaller datasets treat as unobserved.
- Open Access
- 12 August 2025
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 20 (8) , e0329603
Enhancing the gravity model for commuters: Time-and-spatial-structure-based improvements in Japan’s metropolitan areas
Metropolitan commuting flows reveal crucial insights into urban spatial dynamics; however, existing mobility models often struggle to capture the complex, heterogeneous patterns within these regions. This study introduces the Spatially Segregated Urban Gravity (SSUG) model, a novel approach that synergistically combines urban classification with gravity-based flow prediction to address this limitation. The SSUG model’s key innovations include: (1) demonstrating the existence of different scaling laws in metropolitan areas, (2) identifying the existence of data-driven bifurcation that delineates urban-suburban commuting behaviors, (3) employing scaling exponents to reveal spatial segregation, and (4) leveraging high-resolution Global Positioning System (GPS) data for precise deterrence factor measurement. This multifaceted approach enables simultaneous improvement in flow prediction accuracy and robust urban functional classification. Empirical validation across six diverse Japanese metropolitan areas—Tokyo, Osaka, Nagoya, Fukuoka, Sendai, and Sapporo—demonstrates the SSUG model’s superior predictive power compared to traditional gravity models. Our results unveil previously undetected patterns of spatial structure and functional segregation, particularly highlighting the divergent commuting dynamics between urban cores and suburban peripheries. The SSUG model’s capacity to discern fine-grained urban-suburban differences while accurately forecasting commuting flows offers transformative potential for evidence-based urban planning. By providing a more nuanced understanding of metropolitan mobility patterns, this study equips policymakers with a powerful tool for optimizing transportation networks, refining land-use strategies, and fostering sustainable urban development in increasingly complex metropolitan landscapes.
- 12 August 2025
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 20 (8) , e0329603
Open AccessEnhancing the gravity model for commuters: Time-and-spatial-structure-based improvements in Japan’s metropolitan areas
Metropolitan commuting flows reveal crucial insights into urban spatial dynamics; however, existing mobility models often struggle to capture the complex, heterogeneous patterns within these regions. This study introduces the Spatially Segregated Urban Gravity (SSUG) model, a novel approach that synergistically combines urban classification with gravity-based flow prediction to address this limitation. The SSUG model’s key innovations include: (1) demonstrating the existence of different scaling laws in metropolitan areas, (2) identifying the existence of data-driven bifurcation that delineates urban-suburban commuting behaviors, (3) employing scaling exponents to reveal spatial segregation, and (4) leveraging high-resolution Global Positioning System (GPS) data for precise deterrence factor measurement. This multifaceted approach enables simultaneous improvement in flow prediction accuracy and robust urban functional classification. Empirical validation across six diverse Japanese metropolitan areas—Tokyo, Osaka, Nagoya, Fukuoka, Sendai, and Sapporo—demonstrates the SSUG model’s superior predictive power compared to traditional gravity models. Our results unveil previously undetected patterns of spatial structure and functional segregation, particularly highlighting the divergent commuting dynamics between urban cores and suburban peripheries. The SSUG model’s capacity to discern fine-grained urban-suburban differences while accurately forecasting commuting flows offers transformative potential for evidence-based urban planning. By providing a more nuanced understanding of metropolitan mobility patterns, this study equips policymakers with a powerful tool for optimizing transportation networks, refining land-use strategies, and fostering sustainable urban development in increasingly complex metropolitan landscapes.