Conditional Value-at-Risk for Log-Distributions

Khokhlov, Valentyn

doi:10.2139/ssrn.3197929

Cited by 5 publications

(5 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…2 ) [24]. For general distribution, CVaR can be estimated from sampling, which is computationally expensive.…”

Section: B Conditional Value At Riskmentioning

confidence: 99%

TRC: Trust Region Conditional Value at Risk for Safe Reinforcement Learning

Kim

2022

IEEE Robot. Autom. Lett.

View full text Add to dashboard Cite

As safety is of paramount importance in robotics, reinforcement learning that reflects safety, called safe RL, has been studied extensively. In safe RL, we aim to find a policy which maximizes the desired return while satisfying the defined safety constraints. There are various types of constraints, among which constraints on conditional value at risk (CVaR) effectively lower the probability of failures caused by high costs since CVaR is a conditional expectation obtained above a certain percentile. In this paper, we propose a trust region-based safe RL method with CVaR constraints, called TRC. We first derive the upper bound on CVaR and then approximate the upper bound in a differentiable form in a trust region. Using this approximation, a subproblem to get policy gradients is formulated, and policies are trained by iteratively solving the subproblem. TRC is evaluated through safe navigation tasks in simulations with various robots and a sim-toreal environment with a Jackal robot from Clearpath. Compared to other safe RL methods, the performance is improved by 1.93 times while the constraints are satisfied in all experiments.

show abstract

“…2 ) [24]. For general distribution, CVaR can be estimated from sampling, which is computationally expensive.…”

Section: B Conditional Value At Riskmentioning

confidence: 99%

TRC: Trust Region Conditional Value at Risk for Safe Reinforcement Learning

Kim

2022

IEEE Robot. Autom. Lett.

View full text Add to dashboard Cite

show abstract

“…The hazard in that situation (that is, with r = 1) is Where . A popular statistical concept in the business world (Khokhlov, 2016 for details), Tail Value at Risk (Tar) is Similarly, all the Bayesian results for the binomial random variable, y 3 are easily derivable by interchanging γ and δ in all above expressions. The survival function of the random number, Y 3 with symptomatic symptoms is The hazard rate, h ( y )for the symptomatic sign is The binomial distribution of those with symptomatic signs has a finite memory confirming that the usual binomial probability trend of those with symptomatic signs does possess a finite memory.…”

Section: Poisson Heterogeneity: Derivationsmentioning

confidence: 99%

“…Where . A popular statistical concept in the business world (Khokhlov, 2016 for details), Tail Value at Risk (Tar) is …”

Section: Poisson Heterogeneity: Derivationsmentioning

confidence: 99%

Heterogeneity versus the COVID-19 Pandemic

Shanmugam

Ledlow

Singh

2021

Preprint

View full text Add to dashboard Cite

In this paper, heterogeneity is formally defined, and its properties are explored. We define and distinguish observable versus non-observable heterogeneity. It is proposed that heterogeneity among the vulnerable is a significant factor in the contagion impact of COVID-19, as demonstrated with incidence rates on a Diamond Princess Cruise ship in February 2020. Given the nature of the disease, its heterogeneity and human social norms, pre-voyage and post-voyage quick testing procedures may become the new standard for cruise ship passengers and crew. The technological advances in testing available today would facilitate more humanistic treatment as compared to more archaic quarantine and isolation practices for all onboard ship. With quick testing, identification of those infected and thus not allowed to embark on a cruise or quarantining those disembarking and other mitigation strategies, the popular cruise adventure could be available safely again. Whatever the procedures implemented, the methodological purpose of this study should add valuable insight in the modeling of disease and specifically, the COVID-19 virus.

show abstract

“…The tail value at risk (TVaR) was created in financial discussions to describe the gain or loss in trading stocks over a period. The TVaR is an index to quantify the expected value of a loss given that an event outside a given probability has been noticed (see Khokhlov, 2016 for details). Such concept is useful to interpret the spread of corona virus as well.…”

Section: Var Y S H E Y H E S H Var Y H E Y H Var S H E Y H Var Y H Var S H E S H E Y Hmentioning

confidence: 99%

“…Using expressions (0.3) and (0.4), we estimate (see Table 1) the incidence rate and restriction level parameters (namely, Θ and β ) for each region. The largest incidence [12] rate was in Western Pacific region for those exposed to the virus in China. Ironically, the largest restriction level was in South East Asian region for those exposed to the virus in China.…”

Section: Var Y S H E Y H E S H Var Y H E Y H Var S H E Y H Var Y H Var S H E S H E Y Hmentioning

confidence: 99%

Untitled

2020

ABBA

View full text Add to dashboard Cite

In this article, we untangle the mystic of globally scaring corona virus (also, recognized as COVID 19). For this purpose, we define and utilize the index of dispersion to construct a new methodology based on the corona incidence rate, θ and a restriction parameter, β via an incidence rate restricted Poisson (IRRP) model for the data analysis. The IRRP model was introduced by Shanmugam [1] for another purpose and it is found to be quite suitable to understand the mystic nature of non-quantifiable restriction(s) imposed on the corona exposure/treatment in China and everywhere in the world as it appears. The publicly available [2] corona data as of 27 February 2020 in the World Health Organization's (WHO) web page are analyzed and interpreted in this article. In specific, the new methodology detects non-trivial patterns not only in corona incidences but also in the restrictions imposed to deal with the pandemic like this corona virus and this discovery would have been obscured otherwise. The healthcare policies with respect to corona incidences should be congruent to much needed restrictions such preventive hygienic practices as minimal necessities but can go as far as isolation in quarantines as the data evidences enlisted in this article are providing a strong support for them.

show abstract

Conditional Value-at-Risk for Log-Distributions

Cited by 5 publications

References 8 publications

TRC: Trust Region Conditional Value at Risk for Safe Reinforcement Learning

TRC: Trust Region Conditional Value at Risk for Safe Reinforcement Learning

Heterogeneity versus the COVID-19 Pandemic

Untitled

Contact Info

Product

Resources

About