論文 - 山本　俊行

Driver's route choice behavior: Analysis by data mining algorithms 査読有り

Toshiyuki Yamamoto, Ryuichi Kitamura and Junichiro Fujii

Transportation Research Record   ( 1807 ) 頁： 59-66   2002年

　詳細を見る

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（学術雑誌）  

Decision trees and production rules, which are among the methods used in knowledge discovery and data mining, are applied in this study to investigate drivers' route choice behavior. These methods have an advantage over artificial neural networks, another data mining method often used in the analysis of travel behavior, that they facilitate the determination of the relationships between the explanatory variables and the choice. Specifically, the C4.5 algorithm, which produces a decision tree and a set of production rules from the tree, is applied in this study. Two surveys were carried out to collect data on drivers' route choice behavior between two alternative routes on expressway networks. The two data sets include the expected minimum, maximum and average travel times along each alternative route as indicated by the respondent, as well as his socio-demographic attributes. The results of the analyses suggest that different expected travel times influence route choice in different cases, and that a maximum or average travel time determines route choice in some cases regardless of other attributes. The results of a comparison analysis between the C4.5 algorithm and discrete choice models indicate the superior ability offered by the former in representing drivers' route choice.

A Simulation Analysis for the Management of an Electric-Vehicle Sharing System: The Case of the Kyoto Public-Car System 査読有り

Shoichiro Nakayama, Toshiyuki Yamamoto, and Ryuichi Kitamura

Transportation Research Record   ( 1791 ) 頁： 99-104   2002年

　詳細を見る

記述言語：英語   掲載種別：研究論文（学術雑誌）  

A multi-station electric-vehicle (EV) sharing system has been in operation on an experimental basis in Kyoto, Japan. Members of the system can check out EVs at a station and return them at any station. To reduce the cost of system operation, EVs are not reallocated to stations by the operator. This feature, at the same time, may delimit the efficiency of the system because of the mismatch between the spatial distribution of the demand and the distribution of the EVs among the stations. We construct a simulation model of system operation, and configure the optimal management of the system using genetic algorithms, with the number of checkouts per vehicle set as the objective function to be maximized. The number of vehicles, parking capacity at each station and number of members are among the decision variables. The results suggest that the optimal number of EVs is about a half of the total number of parking stalls. This implies that the efficiency of the system does not necessarily improve as more EVs are introduced. Also, the results suggest that even if a new station is introduced into the system, the efficiency does not necessarily improve.

Modeling the Simultaneity in Injury Causation in Multivehicle Collisions 査読有り

Ouyang, Y., Shankar, V. and Yamamoto, T.

Transportation Research Record   ( 1784 ) 頁： 143-152   2002年

　詳細を見る

記述言語：英語   掲載種別：研究論文（学術雑誌）  

Constrained by the single-equation specification, traditional logit or probit models for traffic accident severities can only model one severity at a time. Single-equation models fails to account for the simultaneity of injury severity outcomes in multivehicle collisions. An attempt to account for this simultaneity is made by presenting a simultaneous logit model of endogenous interrelationships among injury severities in multivehicle collisions. Car-truck collision data in Washington State between 1990 and 1996 are used for a simultaneous binary logit model to determine the effects of various exogenous factors in the presence of injury simultaneity, and empirical comparisons with traditional single-equation logit models are also provided. The methodology presented provides for significant efficiency gain when correlation between injury severities is high and can easily be extended to multiple severity categories in multi-vehicle collisions. Empirically, it is also found that parameter estimates for key roadway design variables may vary greatly between a simultaneous logit and its single-equation counterparts. Another finding from this methodology is that preventive measures for head-on collisions, or enforcement of drunk-driving regulations among truck drivers, as well as limiting the occurrence of high-speed curve designs on truck routes, appear to be the most likely factors for significant potential savings in social costs from car-truck collisions.

路側障害物との接触事故による身体損傷程度の分析

山本俊行，Shankar, V.

交通科学   32 巻   頁： 77-80   2002年

　詳細を見る

担当区分：筆頭著者   記述言語：日本語   掲載種別：研究論文（学術雑誌）  

路面電車の需要予測手法に関する実務的課題と対応例

細川寛，山本俊行，飯田祐三，白水靖郎

土木学会第57回年次学術講演会講演概要集CD-ROM     頁： IV-400   2002年

　詳細を見る

記述言語：日本語  

Equilibrium or Disequilibrium? Behavioral Aspect of Driver's Route Choice

Takayuki Morikawa, Shinya Kurauchi and Toshiyuki Yamamoto

Proceedings of the International Symposium on Transport Simulation, R. Kitamura and M. Kuwahara (eds.)     2002年

　詳細を見る

記述言語：英語  

An Overview of PCATS/DEBNetS Micro-Simulation System: Its Development, Extension, and Application to Demand Forecasting

Ryuichi Kitamura, Akira Kikuchi, Satoshi Fujii and Toshiyuki Yamamoto

Proceedings of the International Symposium on Transport Simulation, R. Kitamura and M. Kuwahara (eds.)     頁： 43-62   2002年

　詳細を見る

記述言語：英語  

Modeling the Simultaneity in Injury Causation in Multi-vehicle Collisions 査読有り

Yanfeng Ouyang, Venkataraman N. Shankar, Toshiyuki Yamamoto

Preprint CD-ROM of the 81th Annual Meeting of the Transportation Research Board     2002年

　詳細を見る

記述言語：英語  

A methodology for weighting data from complex endogenous sampling 査読有り

Ryuichi Kitamura, Toshiyuki Yamamoto and Hiromu Sakai

Preprint CD-ROM of the 81th Annual Meeting of the Transportation Research Board     2002年

　詳細を見る

記述言語：英語  

The methodology for weighting endogenous samples is extended in this study and applied to a multi-dimensional choice-based sample and an enriched sample. It is shown that a unified weight exists for a sample obtained from a complex endogenous sampling scheme. Problems with conventional weighting methods for samples from endogenous sampling, such as roadside surveys, are pointed out and an improved weighting scheme is proposed. Coefficient estimates of discrete choice models with and without weights are compared, and the importance of applying appropriate weights and a valid covariance estimator is pointed out.

A simulation analysis for the management of an electric-vehicle sharing system: the case of the Kyoto Public-Car System 査読有り

Shoichiro Nakayama, Toshiyuki Yamamoto and Ryuichi Kitamura

Preprint CD-ROM of the 81th Annual Meeting of the Transportation Research Board     2002年

　詳細を見る

記述言語：英語  

A multi-station electric-vehicle (EV) sharing system has been in operation on an experimental basis in Kyoto, Japan. Members of the system can check out EVs at a station and return them at any station. To reduce the cost of system operation, EVs are not reallocated to stations by the operator. This feature, at the same time, may delimit the efficiency of the system because of the mismatch between the spatial distribution of the demand and the distribution of the EVs among the stations. We construct a simulation model of system operation, and configure the optimal management of the system using genetic algorithms, with the number of checkouts per vehicle set as the objective function to be maximized. The number of vehicles, parking capacity at each station and number of members are among the decision variables. The results suggest that the optimal number of EVs is about a half of the total number of parking stalls. This implies that the efficiency of the system does not necessarily improve as more EVs are introduced. Also, the results suggest that even if a new station is introduced into the system, the efficiency does not necessarily improve.

An analysis of the effects of French vehicle inspection program and grant for scrappage on household vehicle transaction 査読有り

Toshiyuki Yamamoto, Jean-Loup Madre and Ryuichi Kitamura

Preprint CD-ROM of the 81th Annual Meeting of the Transportation Research Board     2002年

　詳細を見る

担当区分：筆頭著者   記述言語：英語  

French government has implemented a periodical vehicle inspection program, which aims at maintaining proper functioning of the vehicle and ensuring the emissions control systems installed on the vehicle work properly. Also, an incentive program for scrapping old vehicles was introduced in 1994 through 1996 to promote the replacement of those vehicles with higher emissions by newer vehicles with lower emissions. A hazard-based duration model of household vehicle transaction behavior has been developed in this study to examine the effects of the inspection program and the grant for scrappage on vehicle transaction timing. The model is developed as a competing risks model assuming the following three types of competing risks: replacing one of the vehicles in the household fleet, disposing of one vehicle in the fleet, and acquiring one vehicle to add to the fleet. The empirical analysis is carried out using the panel data of French households' vehicle ownership from 1984 to 1998, obtained by the panel survey called Parc-Auto, which has been conducted by a French marketing firm, SOFRES, since 1976. The long panel observation period facilitates the introduction of macro-economic indicators into the model, enabling the analysis to distinguish the effects of policy measures from macro-economic factors. The empirical results indicate that the expected vehicle holding duration becomes 1.23 times longer under the inspection program than before the program commenced, given that the vehicle is replaced by another vehicle at the end of the holding duration; and that the conditional probability of replacing a vehicle aged 10 years and over becomes 1.19 times higher when the grant for scrappage is available.

An analysis of driver's route choice behavior by data mining algorithms 査読有り

Toshiyuki Yamamoto, Ryuichi Kitamura and Junichiro Fujii

Preprint CD-ROM of the 81th Annual Meeting of the Transportation Research Board     2002年

　詳細を見る

担当区分：筆頭著者   記述言語：英語  

Decision trees and production rules, which are among the methods used in knowledge discovery and data mining, are applied in this study to investigate drivers' route choice behavior. These methods have an advantage over artificial neural networks, another data mining method often used in the analysis of travel behavior, that they facilitate the determination of the relationships between the explanatory variables and the choice. Specifically, the C4.5 algorithm, which produces a decision tree and a set of production rules from the tree, is applied in this study. Two surveys were carried out to collect data on drivers' route choice behavior between two alternative routes on expressway networks. The two data sets include the expected minimum, maximum and average travel times along each alternative route as indicated by the respondent, as well as his socio-demographic attributes. The results of the analyses suggest that different expected travel times influence route choice in different cases, and that a maximum or average travel time determines route choice in some cases regardless of other attributes. The results of the comparison analyses between the C4.5 algorithm and discrete choice models indicate the superior ability offered by the algorithm in representing drivers' route choice.

Bivariate ordered-response probit model of driver's and passenger's injury severities in collision with fixed objects 査読有り

Toshiyuki Yamamoto and Venkataraman N. Shankar

Preprint CD-ROM of the 81th Annual Meeting of the Transportation Research Board     2002年

　詳細を見る

担当区分：筆頭著者   記述言語：英語  

Previous studies have examined injury severity of vehicle accidents in relation with human factors, vehicle type, highway design, and environmental conditions. Several types of statistical models of injury severity have been developed based on driver's injury in some studies, and on most severely injured occupant's severity in others. A bivariate ordered-response probit model of driver's and most severely injured passenger's severity in collisions with fixed objects is developed in this study. Exact severity of passenger is not necessarily observed, especially when only most severe injury of the accident and driver's injury are recorded in the police reports. To accommodate passenger severities as well, we explicitly develop a partial observability model of passengers' injury severity in multi-occupant vehicle. The model has consistent coefficients for the driver's severity between single-occupant vehicle and multiple-occupant vehicle accidents, and provides more efficient coefficient estimates by taking into account the common unobserved factors between driver's severity and passenger's severity. The results of the empirical analysis using 4-year statewide accident data in Washington State reveal the effects of driver's characteristics, vehicle attributes, types of objects, and environmental conditions on both driver's and passengers' injury severities, and that their severities have different elasticities to some of the risk factors.

再配車によらない電気自動車の共同利用システムの効率化に関する研究 査読有り

中山晶一朗，山本俊行，北村隆一

土木計画学研究・論文集   19 巻   頁： 481-487   2002年

　詳細を見る

記述言語：日本語   掲載種別：研究論文（学術雑誌）  

環境負荷の小さい電気自動車の普及のために，近年，電気自動車の共同利用が注目されている．複数の駐車場を持つ共同利用システムは，特定の駐車場に車両が集中するなどの問題があり，それを効率的に運営するためには車両・駐車スペースの配置をはじめとする様々な変数を最適な値に設定する必要がある．そこで，本研究では，京都パブリックカーシステムを取り上げ，遺伝的アルゴリズムを用いて諸変数の最適化を行い，その効率的な運営について考察した．結果として，車両数の約2倍弱の駐車スペースが必要であることや新規に駐車場を開設することは必ずしも運営効率を上げるとは限らないことなどが分かった．

In-vehicle route recommendation system with learning by neural networks 査読有り

Toshiyuki Yamamoto, Shoichiro Nakayama and RYuichi Kitamura

Proceedings of Joint 1st International Conference on Soft Computing and Intelligent Systems and 3rd International Symposium on Advanced Intelligent Systems (CD-ROM)     2002年

　詳細を見る

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（学術雑誌）  

A two-stage route recommendation system is developed. The system consists of the minimum cost path search stage and the route choice stage. Several routes are selected according to multiple additive cost functions in the former stage, and a recommendation route is chosen from among the selected routes according to another function capable of nonlinear manipulations in the latter stage. These functions are developed as neural networks. The effectiveness of the system is examined using an experimental data set.

Modelling the Effects of a Periodical Vehicle Inspection Program on Household Vehicle Ownership 査読有り

Toshiyuki Yamamoto, Ryuichi Kitamura and Hiroaki Fujii

Travel Behaviour Research: The Leading Edge     頁： 435-450   2001年

　詳細を見る

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（学術雑誌）  

路側障害物との接触事故による身体損傷程度の分析

山本俊行，Shankar, V.

大阪交通科学研究会平成13年度学術研究発表会講演論文集     2001年

　詳細を見る

記述言語：日本語  

車検制度が世帯の自動車取り替え更新行動に及ぼす影響の分析 査読有り

山本俊行，北村隆一，藤井宏明

土木学会論文集   ( 667/IV-50 ) 頁： 137-146   2001年

　詳細を見る

担当区分：筆頭著者   記述言語：日本語   掲載種別：研究論文（学術雑誌）  

本研究では，我が国の自動車保有行動を規定する大きな要因の一つであると考えられる車検制度が世帯の自動車取り替え更新行動に及ぼす影響を明らかにするために，生存時間モデルを適用した自動車取り替え更新行動モデルを構築し，京都市で得られたパネルデータを用いた実証分析を行った．パラメータ推定の結果より， 保有自動車の買い替えや破棄を行う確率は車検までの期間が6ヶ月を切ると急激に増加することが示された．また，推定されたモデルを用いたシミュレーション分析の結果からは，車検費用の引き上げは保有台数に影響を与えることなく買い替えを早める効果がある一方で，車検期間の延長は保有自動車の破棄を遅らせ，保有台数の増加を招く可能性のあることが示された．

複雑な内生抽出法に基づく標本への離散選択モデルの適用 査読有り

北村隆一，酒井弘，山本俊行

土木学会論文集   ( 667/IV-50 ) 頁： 103-111   2001年

　詳細を見る

記述言語：日本語   掲載種別：研究論文（学術雑誌）  

本論文では内生抽出標本の重み付け理論を拡張し，加重層別標本抽出法や多次元選択肢別標本抽出法による標本への適用を論じる．これらの複雑な内生抽出標本法により得られた標本に適用可能な一意的な重みが存在することを示し，従来一般的になされてきた重み付け法の問題点を指摘するとともに，その改良を提案する．さらに，重み付きと重み無しで離散選択モデルを推定した場合の係数値とt-値を比較し，共分散行列を適切に推定することの重要性を示す．

業務トリップチェインにおける経路・出発時刻選択行動の分析 査読有り

山本俊行，北村隆一，熊田善亮

土木学会論文集   ( 667/IV-50 ) 頁： 33-40   2001年

　詳細を見る

担当区分：筆頭著者   記述言語：日本語   掲載種別：研究論文（学術雑誌）  

業務トリップは，通勤トリップと異なり広い時間帯において発生することや，複数の目的地を持つトリップチェインを形成する傾向が強い等の特徴を持つ．本研究では，トリップチェイン前後の業務活動も含めたトリップチェイン全体の費用最小化行動を仮定した，業務トリップの経路・出発時刻選択モデルを構築した．出発時刻選択については最適な出発時刻で総費用が極小値をとるとの条件を用い，連続時間軸上での時刻選択行動をモデル化した．路側配布によるアンケート調査で得られた実際のトリップチェインデータ，および時間帯別所要時間データに基づき未知パラメータを推定した．さらに，出発時刻や経路，立ち回り順序の変更や所要時間の変化に伴なう費用の変化を算出し，モデルの挙動を確認した．