Data-driven prediction of EVAR with confidence in time-varying datasets

Axelrod, Allan; Carlone, Luca; Chowdhary, Girish; Karaman, Sertaç

doi:10.1109/cdc.2016.7799166

Cited by 5 publications

(4 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Norms and Banach spaces induced by the EVaR are studied by Ahmadi-Javid and Pichler (2017). The EVaR is applied in different fields, such as machine learning (Ahmadi-Javid, 2012d), learning-based autonomous systems (Axelrod et al, 2016), approximate statistical models (Watson & Holmes, 2016), and distributionally robust optimization (Postek et al, 2016). When using the model-based approach to represent input data, portfolio optimization with the EVaR can be done efficiently for a broad range of return rates with known distributional properties; for example, when return rates follow arbitrary independent distributions, or when they follow a generalized linear multi-factor model, inspired by Sharp (1963).…”

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Portfolio optimization with entropic value-at-risk

Ahmadi‐Javid

Fallah-Tafti

2019

European Journal of Operational Research

View full text Add to dashboard Cite

The entropic value-at-risk (EVaR) is a new coherent risk measure, which is an upper bound for both the value-at-risk (VaR) and conditional value-at-risk (CVaR). As important properties, the EVaR is strongly monotone over its domain and strictly monotone over a broad sub-domain including all continuous distributions, while well-known monotone risk measures, such as VaR and CVaR lack these properties. A key feature for a risk measure, besides its financial properties, is its applicability in large-scale sample-based portfolio optimization. If the negative return of an investment portfolio is a differentiable convex function, the portfolio optimization with the EVaR results in a differentiable convex program whose number of variables and constraints is independent of the sample size, which is not the case for the VaR and CVaR. This enables us to design an efficient algorithm using differentiable convex optimization. Our extensive numerical study shows the high efficiency of the algorithm in large scales, compared to the existing convex optimization software packages. The computational efficiency of the EVaR portfolio optimization approach is also compared with that of CVaR-based portfolio optimization. This comparison shows that the EVaR approach generally performs similarly, and it outperforms as the sample size increases. Moreover, the comparison of the portfolios obtained for a real case by the EVaR and CVaR approaches shows that the EVaR approach can find portfolios with better expectations and VaR values at high confidence levels.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Portfolio optimization with entropic value-at-risk

Ahmadi‐Javid

Fallah-Tafti

2019

European Journal of Operational Research

View full text Add to dashboard Cite

show abstract

“…CRMs were proposed within the operations research community and have played an influential role within the modern theory of risk in finance (Acerbi, 2002; Acerbi and Tasche, 2002; Rockafellar, 2007; Rockafellar and Uryasev, 2000). This theory has also recently been adopted for risk-sensitive (RS) model predictive control and decision making (Chow and Pavone, 2014; Chow et al, 2015), and guiding autonomous robot exploration for maximizing information gain in time-varying environments (Axelrod et al, 2016).…”

Section: Introductionmentioning

confidence: 99%

Risk-sensitive inverse reinforcement learning via semi- and non-parametric methods

Singh

Lacotte

Majumdar

et al. 2018

The International Journal of Robotics Research

View full text Add to dashboard Cite

The literature on Inverse Reinforcement Learning (IRL) typically assumes that humans take actions in order to minimize the expected value of a cost function, i.e., that humans are risk neutral. Yet, in practice, humans are often far from being risk neutral. To fill this gap, the objective of this paper is to devise a framework for risk-sensitive IRL in order to explicitly account for a human's risk sensitivity. To this end, we propose a flexible class of models based on coherent risk measures, which allow us to capture an entire spectrum of risk preferences from risk-neutral to worst-case. We propose efficient non-parametric algorithms based on linear programming and semi-parametric algorithms based on maximum likelihood for inferring a human's underlying risk measure and cost function for a rich class of static and dynamic decision-making settings. The resulting approach is demonstrated on a simulated driving game with ten human participants. Our method is able to infer and mimic a wide range of qualitatively different driving styles from highly risk-averse to risk-neutral in a data-efficient manner. Moreover, comparisons of the Risk-Sensitive (RS) IRL approach with a risk-neutral model show that the RS-IRL framework more accurately captures observed participant behavior both qualitatively and quantitatively, especially in scenarios where catastrophic outcomes such as collisions can occur. *

show abstract

“…These metrics were introduced in the operations research literature and have played an influential role within the modern theory of risk in finance [40,4,3,39]. This theory has also recently been adopted for risk-sensitive (RS) Model Predictive Control and decision making [12,13], and autonomous exploration [8].…”

Section: Introductionmentioning

confidence: 99%

Risk-sensitive Inverse Reinforcement Learning via Coherent Risk Models

Majumdar¹,

Singh²,

Mandlekar³

et al. 2017

Robotics: Science and Systems XIII

View full text Add to dashboard Cite

Abstract-The literature on Inverse Reinforcement Learning (IRL) typically assumes that humans take actions in order to minimize the expected value of a cost function, i.e., that humans are risk neutral. Yet, in practice, humans are often far from being risk neutral. To fill this gap, the objective of this paper is to devise a framework for risk-sensitive IRL in order to explicitly account for an expert's risk sensitivity. To this end, we propose a flexible class of models based on coherent risk metrics, which allow us to capture an entire spectrum of risk preferences from risk-neutral to worst-case. We propose efficient algorithms based on Linear Programming for inferring an expert's underlying risk metric and cost function for a rich class of static and dynamic decision-making settings. The resulting approach is demonstrated on a simulated driving game with ten human participants. Our method is able to infer and mimic a wide range of qualitatively different driving styles from highly risk-averse to risk-neutral in a data-efficient manner. Moreover, comparisons of the RiskSensitive (RS) IRL approach with a risk-neutral model show that the RS-IRL framework more accurately captures observed participant behavior both qualitatively and quantitatively.

show abstract

Data-driven prediction of EVAR with confidence in time-varying datasets

Cited by 5 publications

References 10 publications

Portfolio optimization with entropic value-at-risk

Portfolio optimization with entropic value-at-risk

Risk-sensitive inverse reinforcement learning via semi- and non-parametric methods

Risk-sensitive Inverse Reinforcement Learning via Coherent Risk Models

Contact Info

Product

Resources

About