A5101831407- Transportation and Mobility Innovations
- Transportation Planning and Optimization
- Traffic control and management
- Sharing Economy and Platforms
- Smart Parking Systems Research
- Economic theories and models
- Urban and Freight Transport Logistics
- Corporate Finance and Governance
- Urban Transport and Accessibility
- Advanced Bandit Algorithms Research
- Auction Theory and Applications
- Evacuation and Crowd Dynamics
- Electric Vehicles and Infrastructure
- Energy, Environment, and Transportation Policies
- Speech Recognition and Synthesis
- Private Equity and Venture Capital
- Robotic Path Planning Algorithms
- Indoor and Outdoor Localization Technologies
- Network Security and Intrusion Detection
- Higher Education and Teaching Methods
- Traffic Prediction and Management Techniques
- Optimization and Search Problems
- Power Line Communications and Noise
- COVID-19 epidemiological studies
- Financial Markets and Investment Strategies
Clemson University
2016-2024
University of Iowa
2024
Guangzhou Regenerative Medicine and Health Guangdong Laboratory
2024
Amazon (United States)
2023
North China University of Technology
2022
Sichuan University of Science and Engineering
2022
University of Michigan
2016-2021
Institute of High Energy Physics
2021
Chinese Academy of Sciences
2021
Cornell University
2021
We consider a periodic-review single-product inventory system with fixed cost under censored demand. Under full demand distributional information, it is well known that the celebrated (s, S) policy optimal. In this paper, we assume firm does not know distribution priori and makes adaptive ordering decisions in each period based only on past sales (a.k.a. demand). Our performance measure regret, which difference between feasible learning algorithm clairvoyant (full-information) benchmark....
Abstract The immunoprotective components control COVID-19 disease severity, as well long-term adaptive immunity maintenance and subsequent reinfection risk discrepancies across initial remain unclarified. Here, we longitudinally analyzed SARS-CoV-2-specific immune effectors during the acute infection convalescent phases of 165 patients with categorized by severity. We found that early robust CD4 + CD8 T cell responses ameliorate progression shortened hospital stay, while delayed attenuated...
This paper presents evidence on cash holdings for Japanese firms listed the Tokyo Stock Exchange, focusing impact of corporate governance factors in and implication to firm value. We find that insider ownership bank relations play a significant role determining holdings. Our results indicate foreign stockholders select profitable invest, these have higher levels cash. document lead agency problems value negatively, characteristics affect negative relation between
We propose a neural language modeling system based on low-rank adaptation (LoRA) for speech recognition output rescoring. Although pretrained models (LMs) like BERT have shown superior performance in second-pass rescoring, the high computational cost of scaling up pretraining stage and adapting to specific domains limit their practical use Here we present method decomposition train rescoring model adapt it new using only fraction (0.08%) parameters. These inserted matrices are optimized...
The ever-growing quantity of audio-visual information accessible today may be effectively managed by recommender systems, which assist users in discovering new and interesting topics. An increasing number customized suggestion apps have emerged on the World Wide Web last decade. Recommendation systems can’t function without precise behaviour modelling users. conventional wisdom about friend algorithms leaves out crucial user data, leading to a misleading portrayal their actions. common...
Mobility-on-Demand Transit (MoDT) is a suitable solution for linking packed urban centers to low-demand suburban areas. Meanwhile, micromobility services, including dockless bikesharing and electric scooters, are growing exponentially worldwide, providing low-cost, low-emission travel mode short home-based trips. We propose an intermodal network in which travelers use the first-/last-mile connections MoDT. The optimal design of formulated as two-stage stochastic program with...
Ride-pooling, which accommodates multiple passenger requests in a single trip, has the potential to substantially enhance throughput of mobility-on-demand (MoD) systems. This paper investigates MoD systems that operate mixed fleets composed “basic supply” and “augmented vehicles. When basic supply is insufficient satisfy demand, augmented vehicles can be repositioned serve rides at higher operational cost. We formulate joint vehicle repositioning ride-pooling assignment problem as two-stage...
We investigate the dynamic pricing problem in on-demand ride-sharing using a continuous-time continuous-space approach. A platform controls two sides of market, supply (vacant vehicles) and demand (customers' trip requests) via setting prices commissions with objective maximizing its expected revenue infinite horizon. The model is described by multi-population traffic flow an intergroup transfer (a system hyperbolic stochastic partial differential equations); subjects to independent...
<p style='text-indent:20px;'>During the Covid-19 pandemic a key role is played by vaccination to combat virus. There are many possible policies for prioritizing vaccines, and different criteria optimization: minimize death, time herd immunity, functioning of health system. Using an age-structured population compartmental finite-dimensional optimal control model, our results suggest that eldest youngest policy deaths. Our model includes infection vaccinated populations. We apply...
End-to-End (E2E) automatic speech recognition (ASR) systems used in voice assistants often have difficulties recognizing infrequent words personalized to the user, such as names and places. Rare non-trivial pronunciations, cases, human knowledge form of a pronunciation lexicon can be useful. We propose PROnunCiation-aware conTextual adaptER (PROCTER) that dynamically injects into an RNN-T model by adding phonemic embedding along with textual embedding. The experimental results show proposed...
Motivated by the meteoric rise in adoption of both ride-hailing services (DiDi, Uber, Lyft, etc.) and dockless bikesharing (Ofo, Mobike, LimeBike, etc.), we propose a multimodal system where passengers ride bikeshare to/from hubs they switch modes carpool. The proposed mutlimodal is generalization existing Uber ExpressPool service. goal this paper to test empirically feasibility system. We accomplish with aid time-stamped taxi origin/destination data from New York City. analysis has two...
We study the problem of jointly pricing and designing a smart transit system, where agency (the platform) controls fleet demand-responsive vehicles (cars) fixed line service (buses). The platform offers commuters menu options (modes) to travel between origin destination (e.g., direct car trip, bus ride, or combination two), make utility-maximizing choice within this menu, given price each mode. goal is determine an optimal set modes display commuters, prices for these modes, design network...
This paper tests two views of bank’s role for Japanese firms. The are confronted with the empirical evidence, allowing them to compete explain firm’s cash holding decisions and implication holdings on firm value. We find that firms closer bank relations hold less cash, but some over‐borrowing. Our results show banks do not monitor their client unlikely push managers take efficient actions maximizing discover cause more severe agency conflicts who have banks.
Real-time parking occupancy information is valuable for guiding drivers' searching spaces. Recently many detection systems using on-vehicle range-based sensors are invented. However, they disregard the practical difficulty of obtaining access to raw sensory data which required feature-based algorithms. In this paper, we focus on transforming short-range radars (SRR) embedded in Advanced Driver Assistance System (ADAS) collect information, and broadcast it through a connected vehicle network....