Toward Trustworthy Decision-Making for Autonomous Vehicles:A Robust Reinforcement Learning Approach with Safety Guarantees

While autonomous vehicles are vital components of intelligent transportation systems,ensuring the trustworthiness of decision-making remains a substantial challenge in realizing autonomous driving.Therefore,we present a novel robust reinforcement learning approach with safety guarantees to attain trustworthy decision-making for autonomous vehicles.The proposed technique ensures decision trustworthiness in terms of policy robustness and collision safety.Specifically,an adversary model is learned online to simulate the worst-case uncertainty by approximating the optimal adversarial perturbations on the observed states and environmental dynamics.In addition,an adversarial robust actor-critic algorithm is developed to enable the agent to learn robust policies against perturbations in observations and dynamics.Moreover,we devise a safety mask to guarantee the collision safety of the autonomous driving agent during both the training and testing processes using an interpretable knowledge model known as the Responsibility-Sensitive Safety Model.Finally,the proposed approach is evaluated through both simulations and experiments.These results indicate that the autonomous driving agent can make trustworthy decisions and drastically reduce the number of collisions through robust safety policies.

AN ADAPTIVE AQM TO PROVIDE DELAY GUARANTEES FOR DIFFERENTIATED SERVICES

RIO(RED with IN and OUT) is the primary queue management mechanism proposed for assured forwarding in the DiffServ (Differentiated Service) framework. Although RIO can generally provide bandwidth guarantees, its queuing delay is sensitive to the traffic load. This paper presents a qualitative explanation for its origin. As a solution, an Adaptive RIO for Delay (ARIO-D) is proposed to provide guaranteed delay for multimedia traffic. Simulation results show that by trading loss for delay, ARIO-D can effectively improve the robustness of RIO under different and dynamic traffic, and provide stable and differentiated performance of queuing delay without any degradation in performance of throughput.

Modeling Policyholder Behavior through Insurance Resonant Marts for Pricing Options and Guarantees

Background: The 11th Annual Conference of Asia-Pacific Risk and Insurance Association was held on July 22-25, 2007 in Taipei, Taiwan. The first author participated in this annual conference where he met the second author who was invited to deliver two plenary speeches on Corporate Governance and Financial Institution Regulation [1] and Alternative Investments for Financial Institutions [2]. The first author was then working as consultant with i-flex solutions, a subsidiary of Oracle and the second author was Vice President of Strategic Business Initiatives Units at ING Life Insurance in its Taiwan operation. The two authors decided to start collaborating on a research paper titled “Modeling Policyholder Behavior through Insurance Resonant Marts for Pricing Options and Guarantees.” The first version of the paper was submitted for research purposes to ING Insurance Risk Management Global Conference 2007 [3] which was held in Beijing, China. Although it was neither presented nor published, the working draft was constantly updated and revised. In 2015, after eight years of continuous research collaboration, the two authors decided to submit the final version of the paper to the 5th World Congress on Engineering and Technology for scholarly presentation. Aim: The competition in the insurance industry is extremely fierce. Insurance companies are under tremendous pressure to retain and increase their customer base, to offer services at attractive rates and provide returns competitive with mutual funds, equities and banks, to achieve profitability across various lines of insurance, to comply with statutory norms etc. Despite having the best of breeds, such as accountants, actuaries [4], lawyers, underwriters, IT experts, consultants, etc., many insurance companies face severe problems to cope with and survive under such pressures. Insurance companies are now striving towards creating innovative products that can match the expectation of the customers with respect to investment returns and risk coverage at competitive rates, which is a very challenging task. Also it is very important to measure the expectations of the customers keeping in mind that those customers are already owners of other financial products. Pricing always follows the expectations and without proper data support, Model risk is imminent. Even if a product is correctly priced, without understanding the behavior of the policyholder towards various financial products will lead to heavy lapses [4]. The authors describe a new framework called UIRDM Approach (Unified Insurance Resonant Data Mart) for the insurance companies wherein this approach stresses the need to think beyond the insurance boundaries.

Improving China's Legal System for Product Quality Guarantees ——A legal analysi s of product quality guarantee regulation by the Government

It is the responsibility of governments to supervise and regulateproduct quality to protect consumers and prevent failures in the product qualityguarantee system.Appropriate supervision will help safeguard the interests ofconsumers,promote product quality,and resolve disputes.The extent of supervisionshould be guided by the principle of need for state intervention andneeds to be highly diversified.

China's Employment Promotion Law Guarantees Right to Employment

China＇s legislature is soliciting comments and suggestions to the draft of Employment Promotion Law as an endeavor to democratize its process of legislation. Up to April 25 the National People＇s Congress had received over 11,000 items of comments and suggestions.

“One Country,Two Systems”Guarantees Hong Kong's Prosperity

Deng Xiaoping met with Margaret Thatcher after signing the Sino-British Joint Declaration in Beijing,stressing that China will adopt a "one country,two systems"policy when it resumes sovereing right over Hong Kong.(19 December 1984)

Price-Matching Guarantees and Entry Deterrence under Incomplete Information

Abstract This paper studies price-matching guarantees in a market where entrant does not have perfect information about incumbent＇s cost. The low-cost incumbent can adopt price-matching guarantees as a signal to distinguish itself from the high-cost type and thus effectively deter entry. On the other hand, the high-cost incumbent can successfully fool the potential entrant under certain conditions. Compared with the equilibriums in situations where the option of offering a price guarantee is not available, the use of this instrument either makes it easier for the low-cost incumbent to signal its cost, or expands the range of parameters over which the high-cost incumbent is able to deter entry successfully.

Bank’s repayment obligations under refund guarantees in shipbuilding finance

The refund guarantee is undoubtedly a significant document required in the Shipbuilding projects. Due to the lack of uniform test for identifying the independent guarantee in different jurisdictions,slight differences to the terms of guarantee coupled with the different facts of each case can result in dramatically different rulings as to the nature of the refund guarantee and bank’s obligation for repayment of installments thereunder. It is also noteworthy that different refund guarantee may provide the beneficiary with different coverage. From the comparison among CSTC Form,Shanghai Form and NEWBUILDCON Form,it is suggested that the terms of the contract and refund guarantee of the NEWBUILDCON are more favorable to the buyers and buyers’ banks than the other two forms.

ON ACTIVATING A CREDITOR-DRIVEN DETERRENCE MECHANISM IN RELATED-PARTY ULTRA VIRES CORPORATE GUARANTEES

I. INTRODUCTION The US Supreme Court Justice Oliver Wendell Holmes once said,‘Pretty much all law consists in forbidding men to do something that they want to do’. Such is the case with laws regulating related party ultra vires corporate guarantees in China, which account for the majority of corporate guarantee cases and continue to grow despite legislative and judiciary deterrence efforts.

The impact of the Rural Minimum Living Standard Guarantee(Rural Dibao)Program on child nutrition outcomes

The Rural Minimum Living Standard Guarantee(Rural Dibao)is an important unconditional cash transfer program to alleviate poverty in rural China.Despite the importance of children’s nutrition in breaking poverty cycles,little is known about the impact of Rural Dibao on child nutrition outcomes.Using China Family Panel Studies(CFPS),this paper examines the effects of Rural Dibao on child nutrition outcomes and investigates potential pathways and heterogeneous effects.We exploit propensity score matching and difference-in-differences techniques to evaluate the effects of the Rural Dibao program on child nutrition outcomes.Our results suggest that Rural Dibao significantly impacts the nutrition outcomes of children up to 15 years of age.Specifically,our results suggest that Rural Dibao improves child height-to-age z-scores by 1.05 standard deviations and lowers the probability of stunting by 11.9 percentage points.Additional analyses suggest that increased protein intake is the main pathway through which Rural Dibao participation contributes to better nutrition outcomes.We also find that the effect of the program is more pronounced among girls,children who are non-left-behind or live with highly educated mothers,and those from low-income families and poor areas.Our findings suggest that Rural Dibao participation helps improve child nutrition outcomes through improving diet quality.

INSTRUCTIONS FOR AUTHORS

1 Legal requirements The author(s) guarantee(s) that the manuscript will not be published elsewhere in any language without the consent of the copyright holders,that the rights of third parties will not be violated,and that the publisher will not be held legally responsible should there be any claims for compensation.Authors wishing to include figures or text passages that have already been published elsewhere are required to obtain permission from the copyright holder(s) and to include evidence that such permission has been granted when submitting their papers.Any material received without such evidence will be assumed to originate from the authors.

INSTRUCTIONS FOR AUTHORS

1 Legal requirements The author(s)guarantee(s)that the manuscript will not be published elsewhere in any language without the consent of the copyright holders,that the rights of third parties will not be violated,and that the publisher will not be held legally responsible should there be any claims for compensation.

INSTRUCTIONS FOR AUTHORS

1 Legal requirements The author(s)guarantee(s)that the manuscript will not be published elsewhere in any language without the consent of the copyright holders,that the rights of third parties will not be violated,and that the publisher will not be held legally responsible should there be any claims for compensation.Authors wishing to include figures or text passages that have already been published elsewhere are required to obtain permission from the copyright holder(s)and to include evidence that such permission has been granted when submitting their papers.Any material received without such evidence will be assumed to originate from the authors.

INSTRUCTIONS FOR AUTHORS

1 Legal requirements.The author(s)guarantee(s)that the manuscript will not be published elsewhere in any language without the consent of the copyright holders,that the rights of third parties will not be violated,and that the publisher will not be held legally responsible should there be any claims for compensation.

INSTRUCTIONS FOR AUTHORS

1 Legal requirements The author(s)guarantee(s)that the manuscript will not be published elsewhere in any language without the consent of the copyright holders,that the rights of third parties will not be violated,and that the publisher will not be held legally responsible should there be any claims for compensation.

INSTRUCTIONS FOR AUTHORS

Legal requirements The author(s) guarantee(s) that the manuscript will not be published elsewhere in any language without the consent of the copyright holders,that the rights of third parties will not be violated,and that the publisher will not be held legally responsible should there be any claims for compensation.

INSTRUCTIONS FOR AUTHORS

1Legal requirements The author(s)guarantee(s)that the manuscript will not be published elsewhere in any language without the consent of the copyright holders,that the rights of third parties will not be violated,and that the publisher will not be held legally responsible should there be any claims for compensation.

INSTRUCTIONS FOR AUTHORS

1 Legal requirements The author(s)guarantee(s)that the manuscript will not be published elsewhere in any language without the consent of the copyright holders,that the rights of third parties will not be violated,and that the publisher will not be held legally responsible should there be any claims for compensation.Authors wishing to include figures or text passages that have already been published elsewhere are required to obtain permission from the copyright holder(s)and to include evidence that such permission has been granted when submitting their papers.

Joint computation offloading and parallel scheduling to maximize delay-guarantee in cooperative MEC systems

The growing development of the Internet of Things(IoT)is accelerating the emergence and growth of new IoT services and applications,which will result in massive amounts of data being generated,transmitted and pro-cessed in wireless communication networks.Mobile Edge Computing(MEC)is a desired paradigm to timely process the data from IoT for value maximization.In MEC,a number of computing-capable devices are deployed at the network edge near data sources to support edge computing,such that the long network transmission delay in cloud computing paradigm could be avoided.Since an edge device might not always have sufficient resources to process the massive amount of data,computation offloading is significantly important considering the coop-eration among edge devices.However,the dynamic traffic characteristics and heterogeneous computing capa-bilities of edge devices challenge the offloading.In addition,different scheduling schemes might provide different computation delays to the offloaded tasks.Thus,offloading in mobile nodes and scheduling in the MEC server are coupled to determine service delay.This paper seeks to guarantee low delay for computation intensive applica-tions by jointly optimizing the offloading and scheduling in such an MEC system.We propose a Delay-Greedy Computation Offloading(DGCO)algorithm to make offloading decisions for new tasks in distributed computing-enabled mobile devices.A Reinforcement Learning-based Parallel Scheduling(RLPS)algorithm is further designed to schedule offloaded tasks in the multi-core MEC server.With an offloading delay broadcast mechanism,the DGCO and RLPS cooperate to achieve the goal of delay-guarantee-ratio maximization.Finally,the simulation results show that our proposal can bound the end-to-end delay of various tasks.Even under slightly heavy task load,the delay-guarantee-ratio given by DGCO-RLPS can still approximate 95%,while that given by benchmarked algorithms is reduced to intolerable value.The simulation results are demonstrated the effective-ness of DGCO-RLPS for delay guarantee in MEC.