TY - JOUR PY - 2022// TI - Dynamic discrete choice model considering survival of uncertain alternatives JO - Proceedings of the Conference of Japan Society of Traffic Engineers A1 - Kondo, Aiko A1 - Urata, Junji A1 - Hato, Eiji SP - 381 EP - 387 VL - 42 IS - N2 - Actions during times of high uncertainty, such as disasters, take place through structures and processes that are different from normal times because of the uncertainty, and the set of decision makers' options themselves can change significantly. Not only options at the time of decision-making, but also future uncertainty naturally influences moment-to-moment decision-making, but existing research overlooks the impact of choice-set evaluation on preferences. In this research, we constructed a dynamic discrete choice model that can express dynamic and stochastic transition of choice sets. Since the set of options at the time of decision making is formed in a non-compensatory manner and the set of uncertain future options is not determined, the structure treated stochastically and compensatively is described using the residual probability of options. We conducted an empirical analysis applying the proposed model to the choice of place to stay during a heavy rain disaster, and clarified the usefulness of this model. === 不確実性の高い災害時等の行動は、その不確実性ゆえ、平時とは異なる構造や過程を経て行われ、意思決定者の選択肢集合自体が大きく変化しうる。意思決定時の選択肢のみならず、将来の不確実性もまた、時々刻々の意思決定に当然影響するが、既存研究では、選択肢集合評価による選好への影響が看過されている。本研究では、選択肢集合が動的・確率的に遷移することを表現できる動的離散選択モデルを構築した。意思決定時の選択肢集合は非補償的に形成し、不確実な将来の選択肢集合は定まらないため、確率的・補償的に扱う構造を選択肢の残留確率を用いて記述した。提案モデルを、豪雨災害時の滞在場所選択行動に適用した実証分析を行い、本モデルの有用性を明らかにした。
Language: ja
LA - ja SN - 2758-3635 UR - http://dx.doi.org/10.14954/jsteproceeding.42.0_381 ID - ref1 ER -