Masayuki Terada

Masayuki Terada

Auto-generated profile, based on publicly available information

Masayuki Terada

NTT DoCoMo, Inc., Tokyo, Japan

Research Focus & Metrics

Academic Career Length

14Years (2013-Present)

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

Representative Publications

View All Publications
1 August 2013 in Statistical Journal of the IAOS
Verified
Journal Article
Closed access
PDF file
NMNaoki Makita
Naoki Makita
MKMasakazu Kimura
Masakazu Kimura
MTMasayuki Terada
Masayuki Terada
MKMotonari Kobayashi
Motonari Kobayashi
YOYuki Oyabu
Yuki Oyabu
For a long time, only national statistics offices have had the capability of producing small area statistics on population, such as, in Japan, the Grid Square Statistics of Population Census (GPC). With the proliferation of mobile phones in daily life and the significant development of ICT, there emerges a new kind of statistics called the Mobile Spatial Statistics (MSS), which estimates small area population derived from the operational data of a mobile phone network. There are several differences between the two statistics, particularly relating to measuring de jure versus de facto population, and research was conducted to examine the validity of MSS by comparing the two statistics. Deviation rates were calculated to compare the 2010 GPC population with the MSS population before dawn on the 2010 Census day and concluded that, to a certain degree, MSS would be plausible for use in densely populated areas. We also found that MSS highlighted the busy blocks of central Tokyo early in the morning. The research proved the utility of MSS, and at the same time, it confirmed the ongoing usefulness of GPC as the benchmark of population statistics.

12 January 2023 in ACM Transactions on Spatial Algorithms and Systems
Verified
Journal Article
Open access
HAHiroto Akatsuka
Hiroto Akatsuka
MTMasayuki Terada
Masayuki Terada
To utilize a huge amount of observation data based on real-world events, a data assimilation process is needed to estimate the state of the system behind the observed data. The Kalman filter is a very commonly used technique in data assimilation, but it has a problem in terms of practical use from the viewpoint of processing efficiency and estimating the deterioration in precision when applied to particularly large-scale datasets. In this article, we propose a method that simultaneously addresses these problems and demonstrate its usefulness. The proposed method improves the processing efficiency and suppresses the deterioration in estimation precision by introducing correction processes focusing on the non-negative nature and sparseness of data in wavelet space. We show that the proposed method can accurately estimate population dynamics on the basis of an evaluation done using population data generated from cellular networks. In addition, the possibility of wide area abnormality detection using the proposed method is shown from a situation analysis of when Category 5 typhoon Hagibis made landfall in Japan. The proposed method has been deployed in a commercial service to estimate real-time population dynamics in Japan.

30 September 2024 in ACM Transactions on Spatial Algorithms and Systems
Verified
Journal Article
Open access
KOKeiichi Ochiai
Keiichi Ochiai
MTMasayuki Terada
Masayuki Terada
MHMakoto Hanashima
Makoto Hanashima
HSHiroaki Sano
Hiroaki Sano
YUYUICHIRO USUDA
YUICHIRO USUDA
In a disaster situation, local and municipal governments need to distribute relief supplies and provide administrative support to evacuees.Although people are supposed to evacuate to evacuation shelters designated by local governments, some people take refuge at non-designated facilities, called non-designated evacuation shelters , due to unavoidable circumstances such as damage on the access routes to designated evacuation shelters. Upon occurrence of a disaster, therefore, it is necessary for the local governments to quickly find the locations of non-designated evacuation shelters. In this article, we propose a method to detect non-designated evacuation shelters based on autoencoder (AE)-based anomaly detection using real-time population dynamics generated from operation data of cellular phone networks. We assume that reconstruction errors of an AE model include both the errors due to characteristic differences between locations and the errors due to anomalies in population dynamics. Thus, we propose to use the ratio of the reconstruction error before and after the earthquake to determine the threshold of anomaly detection. We evaluate the performance of the proposed method on data from three actual earthquakes in Japan. The evaluation results show that our reconstruction error based approach can achieve better accuracy for the actual disaster data compared to a baseline method that exploits statistical anomaly detection.

2 December 2024
Verified
Conference Paper
Open access
KOKeiichi Ochiai
Keiichi Ochiai
MTMasayuki Terada
Masayuki Terada
As data utilization in organizations is advancing in various fields, insights that data brings will be more diverse when it is sourced through collaboration across different organizations. Federated learning, a machine learning method with distributed data across organizations, with local differential privacy protects privacy by sharing only the model parameters and the information necessary for model update, without having to share the data each organization holds. However, there is a problem with local differential privacy, where the amount of noise increases, leading to the degradation in model accuracy. In this paper, we propose a method of reducing the impact of noise compared to conventional federated learning by leveraging private cross-organizational data collaboration, called Private Cross-aggregation Technology (PCT). PCT combines Private Set Intersection Cardinality, Trusted Execution Environment and Differential Privacy, and outputs a cross-tabulation table that is private from input to output. Our method consists of two steps: (1) creating a private cross-tabulation table using PCT, and (2) training a ML using the private cross-tabulation table. The experiment results showed that (1) the classification accuracy of the proposed method was higher than that of the baseline method in situations where the privacy budget is limited, and (2) the computation time of the proposed method was shorter than that of the baseline method.

13 November 2023
Verified
Conference Paper
Open access
KOKeiichi Ochiai
Keiichi Ochiai
MTMasayuki Terada
Masayuki Terada
HSHiroaki Sano
Hiroaki Sano
MHMakoto Hanashima
Makoto Hanashima
YUYUICHIRO USUDA
YUICHIRO USUDA
It is important for local governments to grasp the evacuation situation upon the occurrence of a large-scale disaster because administrative support can be provided, such as opening temporary accommodation facilities or distributing relief supplies. Early identification of locations where population transitions differ from normal conditions will help local governments to identify areas for disaster response. In this paper, we propose a simple yet effective anomaly detection method for population transition using an autoencoder that leverages periodic residual components using real-time population dynamics generated from operation data of cellular phone networks. We evaluated the effectiveness of our proposed method using data from three real earthquakes in Japan. The results demonstrated that our approach, which leverages periodic residual components, outperforms a baseline method that directly uses population transition. Moreover, the experimental results showed that the standard deviation of recall is smaller than that of the baseline method, indicating that utilizing the residual component has a stabilizing effect on accuracy.

1 January 2026 in IET Intelligent Transport Systems
Verified
Journal Article
Open access
PDF file
HAHiroto Akatsuka
Hiroto Akatsuka
KTKeisuke Toyama
Keisuke Toyama
HHHidetaka Homma
Hidetaka Homma
MTMasayuki Terada
Masayuki Terada
Traffic congestion mitigation is essential for reducing environmental impact and improving quality of life. As constructing new infrastructure is costly, enhancing the operational efficiency of existing road networks is crucial. One promising strategy is to influence driver behaviour by providing congestion predictions before route decisions are made. However, accurate long‐term prediction several hours ahead remains challenging when relying solely on conventional traffic data. This study proposes a population‐demand‐congestion prediction framework that leverages real‐time population data derived from cellular networks to capture latent travel demand. Directly predicting travel time from high‐dimensional and sparse population data results in a nonlinear high‐dimension, low‐sample‐size (HDLSS) problem. To address this, the proposed framework decomposes the task into two stages: estimating traffic demand from population distribution using linear regression under an HDLSS setting, and predicting travel time from the estimated demand using a nonlinear model in a lower‐dimensional space. This decomposition aligns the statistical model with the physical causality of traffic formation and improves estimation stability. Empirical evaluation demonstrates accurate congestion prediction several hours ahead. Furthermore, field experiments on major expressways show that providing predictions influences driver behaviour in up to 89.0% of users, demonstrating the framework's practical potential for real‐world ITS applications.