Recently, the artificial-intelligence-based Werewolf Project has been attracting attention. As the Werewolf Game requires human-like artificial intelligence, it is important for researchers to construct an artificial intelligence that has the abilities to beguile and detect a lie. Therefore, we construct an artificial intelligence that can deduce players' roles and communicate adequately with other players.
Recently, many researchers have studied foreign exchange trading using technical analysis. However, it is difficult to achieve profitability using this technique. Therefore, using Genetic Network Programming, we construct a model that considers the technical index signal strength for devising a profitable trading strategy. Finally, we confirmed the effectiveness of our model using historical data of the exchange market.
DAIHINMIN is a kind of card game that is played in Japan. It is known that there are various local rules. It is common to aim to get rid of it quickly. It is a game played by multiple people, and it is classified as a multiplayer incomplete information game because the opponent's hand is not open. It is always 1 in such a card game. It is difficult to play aiming for rank. Therefore, the purpose of this research is to dynamically estimate the ranking that can be achieved from one's hand and the state of the field.
In recent years, there is an interpretability problem that even experts cannot explain the reasoning process of machine learning. A contest featuring interpretability, "First Knowledge Graph Reasoning Challenge 2018," was held in Tokyo. To learn the sense of word in the novel, a previous study proposed a method that was based on word embedding. However, it did not consider the flow of events. In this study, considering co-occurrence words, we propose a method using the principal components regression analysis of the feature vector of words.
マルチエージェントシミュレーションによって群集の移動を計算する研究が行われてい る.このようなシミュレーションでは各エージェントに経路や帰宅開始時刻などのシナリオを与える ことでその後の群集の移動を計算し帰宅人数の推移などを推定することが可能である.しかしなが ら,様々なシナリオを網羅的にシミュレーションすることは,計算資源の観点から限界がある.この 問題に対して,本研究では各エージェントのシナリオと初期座標,途中までのシミュレーション結果 から,その後の結果を推定するニューラルネットワークを提案する.実験の結果,提案手法は短時間 かつ高精度でシミュレーション結果を推定できることが分かった.この知見は,提案手法によってシ ミュレーションの計算過程の一部を代替し,シミュレーションの実行時間を短縮できることを示して いる.
Each year, about 1.35 million people die as a result of traffic accidents in the world. The 2030 Agenda for Sustainable Development also has set an ambitious target of halving the global number of deaths and injuries from road traffic crashes by 2020 [1]. In Japan, Aichi Prefecture has the most traffic accidents by a prefecture for 16 years until 2018. Aichi Prefecture arose the goal "By 2020, reducing the annual number of traffic fatalities within 24 hours after an accident to no more than 155" in the 10th Aichi Prefecture Traffic Safety Plan released in 2016 [2]. In this paper, we analyze the relationship between the number of accidents and traffic volume in Aichi Prefecture to reduce traffic fatalities. In this analysis used the latest traffic accidents data including 2019. As a result of the analysis, we found some relationship between them. Furthermore, we find the possibility of a rapid increase in the number of traffic accidents when the traffic volume exceeds a certain criteria. We will estimate the criteria near future.
Japanese labor productivity is the lowest in the seven advanced countries(G7) despite the fact that many people are working such long hours with stress as to cause industrial accidents such as mental illness and suicide. Therefore, workstyle reforms that improve productivity while maintaining mental health are urgently needed. In the work style reform, health management based on "Work Attitude" such as stress, work engagement, and workaholism is essential, and it is advisable to measure daily Work Attitude in order to be able to detect the sign of the industrial accident quickly. However, the conventional Work Attitude measurement methods are based on a questionnaire once a year, and then they are not suitable for monitoring. Therefore, in this study, we develop Work Attitude PLR (Personal Life Record) collection platform that continuously measures and records Work Attitude that used to be measured only sporadically based on subjective questionnaires, using multimodal information. In this presentation, I describe the data collection experiments we conducted and their analysis for the construction of the proposed Work Attitude PLR collection platform.
物やサービスの利用体験を最大化するために,利用者の利用場面に即した物・サービスの 設計を行う必要が出てきている.この設計を実現するために利用者のモデル化が必要となっている. 我々はアンケートデータからこのモデル化を行う方法とその実際の研究を行ってきた.その一環とし て,アンケートデータに感性的な設問を取り入れることで使用者の感性を反映したデータを収集す ることや,そのデータに対して確率的なクラスタリングすることで使用者の典型的なタイプを見出 すことを行ってきた.先行研究では,この手段を用いて各々の分野でどのような傾向がみられるかに ついての分析を行ってきたが,本論ではクラスタリングされたセグメントの識別力はどれほどである かの検証を複数のアンケートデータを通じて行った.
予算等の資源を配分を行う際,一部に対して優先的・集中的に配分する "選択と集中” 戦略 が選択される場面は珍しくない.ここで,投資に対して見込める利益率の分布が "べき分布” をとり,か つ,見込める利益率の予測ができない場合には選択をせず,遍く対象に投資する方が全体としての利益 が大きくなることが示されている [野田 19].ただし,予算等の資源は有限ではないため,たとえば申請 があったものについてはすべて予算を支出するとした場合,予算が十分に確保できない可能性が高いほ か,当初から利益が見込めない課題を乱発して自己の資源確保最大化を図るような行動も予想され,現 実的には遍く対象に投資することは難しく,一定のフィルタリングは必要になると考えられる.そこで 本報では投資に対して見込める利益率の分布が "べき分布” の場合に,選択と集中の程度と,個々の課題 の利益率に関する予測精度の関係について,モンテカルロ・シミュレーションを通じて明らかにした.
The Ministry of Health, Labour and Welfare (MHLW) has released the COVID-19 Contact- Confirming Application (COCOA) to prevent the spread of COVID-19. However, it is currently underuti- lized and its effectiveness is questionable. Therefore, we propose the nudge system that promotes people's voluntary installation and aims to increase the utilization of COCOA. The proposed system motivates users to install COCOA by reframing the installation as the acquisition of tickets to experience three attractive events: (1) paint, (2) vote, and (3) fortune slip. In this paper, we propose the nudge system promoting the installation and assume scenarios for the use of the system, taking into account the appropriate environment for each event.
Although the Japanese service industry accounts for 70% of the gross domestic product, labor productivity is low, and improving it is very important for making the Japanese economy better. By clarifying the value required by service stakeholders, the value of the service can be increased, leading to an improvement in the labor productivity of services. In this research, as a means to clarify the value required by stakeholders, we have developed the "AI Touch Rally", which has been studied in the past, to make it more practical, such as making it compatible with the real space and the Internet.
The user's persona, which was physically found by the AI touch rally experiment for actual visitors at the Miraikan, is also used online to build picture-story show content with the same value evaluation structure. At that time, as an online template. We report the results of the prior release of the university anthropomorphic manga character diagnostic app as an example of social application.
人口分布の偏り により 生じる感染過程への影響を統計的に同定し、 それをマルチエー ジェント感染シミュレーションに反映する方法を提案する。 COVID-19のような感染過程が見えに く い感染症に対しては、 人の移動モデルを用いた感染拡大シミ ュレーショ ンが対策立案の重要なツー ルとなる。 これらのシミ ュレーショ ンでは人の分布や移動についてある程度の一様性の過程がある。 一方、 実際の人の空間的分布にはかならず偏りがあり、 携帯端末によるデータでのその偏りを確認で きる。 この偏り は人口密度の偏り として、 感染拡大に大きな影響を与える。 本稿では、 この影響につ いて統計的に分析し、 シミ ュレーショ ンにそれを反映させる方法を提案し、 さらにシミ ュレーショ ン 実験でその効果を検証する。