Article
Computer Science, Artificial Intelligence
Hyunjae Kim, Yookyung Koh, Jinheon Baek, Jaewoo Kang
Summary: This study focuses on exploring the spatial understanding and reasoning ability of neural models, presenting spatial reasoning IQ tests and conducting experiments to evaluate the generalization abilities of models. The analysis of results and factors affecting model generalization provides insights for future research, aiming to encourage further exploration into human-level spatial reasoning.
Article
Computer Science, Information Systems
Enrique Hernandez-Orallo, Antonio Armero-Martinez
Summary: This paper proposes the use of human mobility models to evaluate the temporal and spatial risk of transmission of COVID-19, studying synthetic and simulated models and defining different exposure risks to generate a heat map showing exposure risk in evaluated scenarios, providing an effective means for risk management in public spaces.
Article
Automation & Control Systems
Luca Buoncompagni, Syed Yusha Kareem, Fulvio Mastrogiovanni
Summary: Arianna(+) is a framework for designing network ontologies for online human activity recognition within smart homes. It provides a flexible interface for encoding data within multiple contexts and schedules procedures based on logic reasoning. The focus on small ontology networks enhances intelligibility and reduces computational load.
IEEE TRANSACTIONS ON CYBERNETICS
(2022)
Review
Computer Science, Information Systems
Karim Malik, Colin Robertson, Steven A. Roberts, Tarmo K. Remmel, Jed A. Long
Summary: This paper reviews the use of computer vision models and artificial neural networks in geographical analysis, with a focus on the representation and comparison of spatial patterns. The authors find that scale, which is typically considered a model parameter in computer vision, is a contextual element in geographical research. However, convolutional neural networks (CNNs) are relatively robust to small-scale variations due to their ability to learn multiscale features. Although there are still challenges in parameterizing computer vision models to represent multiscale patterns, a typology of scales can provide a framework for guidelines in a geographic context.
INTERNATIONAL JOURNAL OF GEOGRAPHICAL INFORMATION SCIENCE
(2023)
Article
Economics
Michele Bavaro, Fabrizio Patriarca
Summary: This paper investigates the interaction between two channels of intergenerational transmission of inequalities: parental income and networks. It shows the advantages and limitations of job referrals in the labor market by examining their impact on the opportunity gap between workers with parents of different educational levels.
ECONOMIC MODELLING
(2022)
Article
Economics
Alejandro Builes-Jaramillo, Laura Lotero
Summary: The bicycle sharing system in Medellin is an alternative for sustainable transportation. Analyzing the usage patterns and challenges can provide valuable insights for future planning.
JOURNAL OF TRANSPORT GEOGRAPHY
(2022)
Article
Computer Science, Information Systems
Djibril Mboup, Cherif Diallo, Hocine Cherifi
Summary: Mobility is crucial for understanding human contact networks. Existing synthetic network models have limited success in replicating real-world contact networks, but data generated using the STEPS mobility model shows some similarities, providing a direction for improving synthetic network models.
Article
Chemistry, Multidisciplinary
Lennart Adenaw, Quirin Bachmeier
Summary: Accurate modeling of human mobility demand is crucial in transportation system engineering. This article presents an approach to generate activity-based mobility plans that accurately reproduce real-world mobility demand without the complexity of activity-based methods. The approach combines existing trip-based models and mobility surveys, and the results demonstrate its effectiveness on both a microscopic and a macroscopic level.
APPLIED SCIENCES-BASEL
(2022)
Article
Computer Science, Cybernetics
Tariq Ahmad, Jinsong Wu
Summary: Smart video surveillance plays a significant role in public security by storing and evaluating a large amount of continual stream data and generating warnings for undesirable human activities. To address the challenges of human activity recognition in video surveillance, a lightweight spatial-deep features integration using multilayer GRU (SDIGRU) is introduced. Spatial and deep features are extracted from realistic human activity videos and integrated, and informative features are selected to train a multilayer GRU for learning the temporal dynamics of human activity frames sequence. The experimental results on benchmark datasets demonstrate the effectiveness and efficiency of the proposed method compared to existing state-of-the-art techniques.
IEEE TRANSACTIONS ON COMPUTATIONAL SOCIAL SYSTEMS
(2023)
Article
Biochemistry & Molecular Biology
Ludovic Bellier, Anais Llorens, Deborah T. Marciano, Aysegul Gunduz, Gerwin Schalk, Peter Brunner, Robert T. Knight
Summary: This study analyzed the neural dynamics of music perception by reconstructing a recognizable song from direct neural recordings of 29 patients. The results showed the involvement of specific brain regions and response patterns in music perception and suggested the potential application of music in brain-computer interface technology.
Article
Multidisciplinary Sciences
Matan Fintz, Margarita Osadchy, Uri Hertz
Summary: Deep neural network models have the potential to provide new insights into cognitive processes, but their opaque nature limits their usefulness as a scientific tool. This study proposes using DNN models as exploratory tools and characterizing their operation using explicit models to reveal predictable decision patterns.
SCIENTIFIC REPORTS
(2022)
Article
Multidisciplinary Sciences
Yuhao Huang, Jeffrey B. Wang, Jonathon J. Parker, Rajat Shivacharan, Rayhan A. Lal, Casey H. Halpern
Summary: This study reveals the strong relationship between human brain activity and peripheral metabolism, showing that the central nervous system can decode glucose levels based on intracranial activity. It suggests that the CNS plays a crucial role in maintaining glucose homeostasis by sensing and regulating peripheral glucose levels.
NATURE COMMUNICATIONS
(2023)
Article
Computer Science, Information Systems
Saeed Mohsen, Ahmed Elkaseer, Steffen G. Scholz
Summary: In this study, three models were proposed for human activity classification, with experimental results showing that the hybrid CNN-LSTM model outperformed the individual LSTM and CNN models in classifying human activities.
Article
Computer Science, Hardware & Architecture
Vishnu Vardhan Chetlur, Harpreet S. Dhillon
Summary: This article provides an accessible introduction to the Poisson line Cox process (PLCP) and demonstrates its relevance in the study of various performance metrics in vehicular networks. The versatility of the PLCP model enables the study of path distance characteristics, which have useful applications in transportation networks and the Industrial Internet of Things.
Article
Chemistry, Analytical
Gustavo Aquino, Marly G. F. Costa, Cicero F. F. Costa Filho
Summary: This paper investigates the application of human activity recognition (HAR) in wearables and utilizes deep learning methods for classification. The study demonstrates that visual explanations of model decision making can be generated using accelerometer data. The correlation between HAR and biometric user identification (BUI) tasks is explored, and the use of gradient-weighted class activation mapping (grad-CAM) is proposed for visual explanations.
Article
Automation & Control Systems
Hemant Gehlot, Shreyas Sundaram, Satish Ukkusuri
Summary: The passage discusses finding an optimal control policy for a system where components experience disruptions and their health values decrease over time. Depending on the conditions, the optimal control strategy can target the component with the largest state value or the component with the least state minus deterioration rate.
IEEE TRANSACTIONS ON AUTOMATIC CONTROL
(2022)
Article
Automation & Control Systems
Qi Xin, Rui Fu, Satish V. Ukkusuri
Summary: This study proposes a safe and sub-optimal longitudinal control protocol for CAV platoon with uncertain vehicle dynamics and state constraints. By encoding state constraints and speed trajectory tracking stability condition into control constraints, the control stability and performance of the CAV platoon are improved.
JOURNAL OF THE FRANKLIN INSTITUTE-ENGINEERING AND APPLIED MATHEMATICS
(2022)
Article
Environmental Studies
Xiaowei Chen, Jiawei Xue, Zengxiang Lei, Xinwu Qian, Satish Ukkusuri
Summary: This study proposes a novel model for energy-efficient routing of electric vehicles, which obtains the minimal expected energy consumption paths for multiple origin-destination pairs simultaneously. The proposed algorithms outperform traditional shortest trip time and distance path algorithms in terms of energy savings.
TRANSPORTATION RESEARCH PART D-TRANSPORT AND ENVIRONMENT
(2022)
Article
Engineering, Civil
Shagun Mittal, Takahiro Yabe, Fatima Arroyo Arroyo, Satish Ukkusuri
Summary: Accessibility is crucial for any transportation system, and low accessibility can lead to compromised living conditions, low economic growth, and social inequalities. This study uses novel data sets to assess accessibility and poverty-based inequalities in the Greater Maputo region in Mozambique, revealing disparities in driving, transit, and walking accessibility between poor and rich regions.
TRANSPORTATION RESEARCH RECORD
(2023)
Article
Multidisciplinary Sciences
Sandro M. Reia, P. Suresh C. Rao, Marc Barthelemy, Satish V. Ukkusuri
Summary: A new study reveals that city growth in the U.S. is uneven and concentrated in core areas. Intra-city flows tend to move towards external and low-density counties, contributing to urban sprawl. The study also highlights the significant impact of domestic migration on population growth, surpassing natural demographic growth, and driving the heterogeneity in city growth patterns.
NATURE COMMUNICATIONS
(2022)
Article
Transportation
Shagun Mittal, Takahiro Yabe, Indraneel Kumar, Satish Ukkusuri
Summary: This paper uses the NETS database to analyze the entry of establishments in the U.S.-400 region (mainly in Kansas) over a 20-year period. The study finds significant relationships between pre-existing and new-entrant businesses, indicating their importance in entry decisions. These findings are crucial for understanding freight vehicle flows and guiding corridor utilization, as well as informing policies on regional attraction and growth.
TRANSPORTMETRICA A-TRANSPORT SCIENCE
(2022)
Article
Engineering, Electrical & Electronic
Chenbei Lu, Jiaman Wu, Jingshi Cui, Yanyan Xu, Chenye Wu, Marta C. Gonzalez
Summary: The increasing number of electric vehicles (EVs) presents both opportunities and challenges to the power system. In order to address operational difficulties faced by EV charging stations due to incomplete information on EVs' departure times and preferences, an optimal deadline differentiated dynamic price menu is designed. This menu incentivizes EVs to truthfully reveal their departure times and improve operational efficiency.
IEEE TRANSACTIONS ON SMART GRID
(2023)
Article
Engineering, Civil
Eunhan Ka, Smita Sharma, Satish Ukkusuri
Summary: Lane management plays an important role in alleviating traffic congestion and improving road capacity. This study proposed an analytical framework to assess the impacts and economic effects of lane management on the entire road network. The case study of the Alex Fraser Bridge in Vancouver showed that the contraflow lane with movable median barriers significantly improved traffic flow and generated economic benefits.
JOURNAL OF TRANSPORTATION ENGINEERING PART A-SYSTEMS
(2022)
Article
Transportation Science & Technology
Rezaur Rahman, Samiul Hasan
Summary: This paper presents a novel data-driven approach for predicting evacuation traffic at a network scale. By adopting a transfer learning approach, the model combines non-evacuation period traffic data with additional features related to evacuation traffic demand to improve prediction accuracy. The implemented model performs well in predicting evacuation traffic flow and can be applied over a longer forecasting horizon.
TRANSPORTATION RESEARCH PART C-EMERGING TECHNOLOGIES
(2023)
Article
Automation & Control Systems
Suddhasattwa Das, Shakib Mustavee, Shaurya Agarwal, Samiul Hasan
Summary: This article presents a novel data-driven framework for analyzing quasiperiodically driven dynamical systems. It accurately reconstructs the driving dynamics by computing Koopman eigenfrequencies and then reconstructs the driven system using the same framework. The proposed approach is applied to traffic data and provides insights into generating frequencies, accurately reconstructs queue lengths, and makes stable long-term forecasts.
IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS
(2023)
Article
Geosciences, Multidisciplinary
Tasnuba Binte Jamal, Samiul Hasan
Summary: Reduced human activity due to extreme events has significant negative impacts on productivity, economy, and social wellbeing. Community resilience, defined as the ability of a system to manage shocks and return to a steady state, is important in understanding the impacts of extreme events on population activity. This study analyzed aggregate location data from Facebook during Hurricane Ida to quantify community resilience. The loss in resilience of affected communities is associated with disruptions in physical infrastructures, disaster conditions, and socioeconomic disparities.
INTERNATIONAL JOURNAL OF DISASTER RISK REDUCTION
(2023)
Article
Environmental Sciences
Cristobal Pais, Jose Ramon Gonzalez-Olabarria, Pelagie Elimbi Moudio, Jordi Garcia-Gonzalo, Marta C. Gonzalez, Zuo-Jun Max Shen
Summary: This study presents a framework for classifying fire regimes spatially on a global scale based on historical records from 2000 to 2018. It reveals 15 global pyromes with differences in fire-related metrics and shows how factors such as vegetation, climate, and demographic features can result in specific fire regimes. By processing historical wildfire records and dividing the pyromes into 62 regimes based on spatial aggregation patterns, this study provides a spatial framing of contemporary fire regimes. It expands on existing classification efforts and bridges the gaps between global and regional fire studies.
COMMUNICATIONS EARTH & ENVIRONMENT
(2023)
Article
Environmental Studies
Rajat Verma, Satish V. Ukkusuri
Summary: This study introduces a composite measure of walkability and walking, called 'pednet score', which overcomes the limitations of existing walkability measures. By studying three hypothetical variants of pedestrian networks in three North American cities, the study shows that selecting candidate sidewalk and/or crosswalk segments based on the pednet score can significantly increase walking trips and reduce pedestrian trip distances. The results from marginal benefit curves strongly indicate the usefulness of the pednet score as a measure of link criticality in pedestrian network design.
NPJ URBAN SUSTAINABILITY
(2023)
Article
Engineering, Civil
Rajat Verma, Satish V. Ukkusuri
Summary: This study employed deep learning to detect crosswalks using satellite imagery data and proposed an algorithm to assign the detected crosswalks. The effectiveness of this technique was demonstrated in Washington, D.C. and Los Angeles, CA, and the influence of increasing distance threshold on classification performance was explored.
TRANSPORTATION RESEARCH RECORD
(2023)
Article
Economics
Zengxiang Lei, Satish V. Ukkusuri
Summary: This study proposes a reinforcement learning-based approach for the dynamic pricing problem in ride-hailing systems. By translating the problem into a Markov Decision Process, the existence of a deterministic stationary optimal policy is proven. Using the offline learning algorithm TD3, the optimal pricing policy is learned from historical data and applied to the next time slot. Extensive numerical experiments demonstrate the effectiveness of the proposed algorithm in finding the optimal pricing policy and improving platform profit and service efficiency in both small and large networks.
TRANSPORTATION RESEARCH PART B-METHODOLOGICAL
(2023)
