Liang Hong
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Liang Hong has a total of 5 co-authors. Here are the top co-authors by number of shared publications:
Ryan Martin
North Carolina State University
2 shared publications- Boyi Zhuang
Center for Risk and Insurance Research, The Culverhouse College of Business The University of Alabama Tuscaloosa Alabama United States
1 shared publication - Harris Schlesinger
University of Alabama
1 shared publication - Jun Hu
Oakland University
1 shared publication - Xiufeng Xu
The University of Texas at Dallas
1 shared publication
Latest Publications
- 18 February 2026
- Published by Springer Nature in European Actuarial Journal
- Vol. 16 (1) , 69-85
Conformal prediction of future insurance claims in the regression problem
- 18 February 2026
- Published by Springer Nature in European Actuarial Journal
- Vol. 16 (1) , 69-85
Conformal prediction of future insurance claims in the regression problem
- Open Access
A new strategy for finite-sample valid prediction of future insurance claims in the regression setting
The extant insurance literature demonstrates a paucity of finite-sample valid prediction intervals of future insurance claims in the regression setting. To address this challenge, this article proposes a new strategy that converts a predictive method in the unsupervised iid (independent identically distributed) setting to a predictive method in the regression setting. In particular, it enables an actuary to obtain infinitely many finite-sample valid prediction intervals in the regression setting.
Open AccessA new strategy for finite-sample valid prediction of future insurance claims in the regression setting
The extant insurance literature demonstrates a paucity of finite-sample valid prediction intervals of future insurance claims in the regression setting. To address this challenge, this article proposes a new strategy that converts a predictive method in the unsupervised iid (independent identically distributed) setting to a predictive method in the regression setting. In particular, it enables an actuary to obtain infinitely many finite-sample valid prediction intervals in the regression setting.
- Open Access
The true detection probability versus the subjective detection probability of a uniformly optimal search plan
This article investigates the difference between the true detection probability and the subjective probability of a uniformly optimal search plan. Its main contributions are multi-fold. First, it provides a set of examples to show that, in terms of the true detection probability, the uniformly optimal search plan may or may not be optimal. Secondly, it establishes that the true detection probability of the uniformly optimal search plan based on a composite prior can be less than that of the composite uniformly search plan based on different priors. Next, it argues that an open problem is unsolvable. Finally, it shows that the true detection probability of the uniformly optimal search plan converges to one as the search time approaches infinity.
Open AccessThe true detection probability versus the subjective detection probability of a uniformly optimal search plan
This article investigates the difference between the true detection probability and the subjective probability of a uniformly optimal search plan. Its main contributions are multi-fold. First, it provides a set of examples to show that, in terms of the true detection probability, the uniformly optimal search plan may or may not be optimal. Secondly, it establishes that the true detection probability of the uniformly optimal search plan based on a composite prior can be less than that of the composite uniformly search plan based on different priors. Next, it argues that an open problem is unsolvable. Finally, it shows that the true detection probability of the uniformly optimal search plan converges to one as the search time approaches infinity.
- 28 August 2024
- Published by Springer Nature
- p. 257-269
Instantaneous and Limiting Behavior of an n-Node Blockchain Under Cyber Attacks from a Single Hacker
- 28 August 2024
- Published by Springer Nature
- p. 257-269
Instantaneous and Limiting Behavior of an n-Node Blockchain Under Cyber Attacks from a Single Hacker
- 26 June 2024
- Published by Taylor & Francis in North American Actuarial Journal
- Vol. 28 (4) , 932-933
Author’s Reply to Discussion on “Sample Size Determination for Credibility Estimation”
- 26 June 2024
- Published by Taylor & Francis in North American Actuarial Journal
- Vol. 28 (4) , 932-933
Author’s Reply to Discussion on “Sample Size Determination for Credibility Estimation”
- 24 April 2023
- Published by Taylor & Francis in Mathematics Magazine
- Vol. 96 (3) , 347-348
Infinitely Many New Derivations of the Geometric Expectation
Summary Several derivations of the geometric expectation have been proposed by previous authors. This article gives infinitely many new derivations of the geometric expectation.
- 24 April 2023
- Published by Taylor & Francis in Mathematics Magazine
- Vol. 96 (3) , 347-348
Infinitely Many New Derivations of the Geometric Expectation
Summary Several derivations of the geometric expectation have been proposed by previous authors. This article gives infinitely many new derivations of the geometric expectation.
- Open Access
Whoops! It Happened Again: Demand for Insurance That Covers Multiple Risks
This article studies insurance demand in a two-period framework in which an individual faces risks in both current and future periods. Models for insurance with and without the presence of endogenous saving are both discussed. In contrast to what most literature suggests, when decisions on insurance and saving are made separately, insurance alone does not always unambiguously reduce risk, and decision makers might demand more insurance when there is a positive loading on the premium than when the insurance price is actuarially fair. We compare the demand for insurance in our framework with that in a two-period model where risk is concentrated in the second period and derive the conditions under which these demands differ. We examine the effects of risk aversion and derive the conditions under which a more risk-averse individual demands more or less insurance.
Open AccessWhoops! It Happened Again: Demand for Insurance That Covers Multiple Risks
This article studies insurance demand in a two-period framework in which an individual faces risks in both current and future periods. Models for insurance with and without the presence of endogenous saving are both discussed. In contrast to what most literature suggests, when decisions on insurance and saving are made separately, insurance alone does not always unambiguously reduce risk, and decision makers might demand more insurance when there is a positive loading on the premium than when the insurance price is actuarially fair. We compare the demand for insurance in our framework with that in a two-period model where risk is concentrated in the second period and derive the conditions under which these demands differ. We examine the effects of risk aversion and derive the conditions under which a more risk-averse individual demands more or less insurance.
- Open Access
On several properties of uniformly optimal search plans
The uniformly optimal search plan is a cornerstone of the optimal search theory. It is well-known that when the target distribution is circular normal and the detection function is exponential, the uniformly search plan has several desirable properties. This article establishes that these properties hold for any continuous target distribution. The results provide useful information to the search team when they need to choose a non-circular normal target distribution in a real-world search mission.
Open AccessOn several properties of uniformly optimal search plans
The uniformly optimal search plan is a cornerstone of the optimal search theory. It is well-known that when the target distribution is circular normal and the detection function is exponential, the uniformly search plan has several desirable properties. This article establishes that these properties hold for any continuous target distribution. The results provide useful information to the search team when they need to choose a non-circular normal target distribution in a real-world search mission.
- 2 October 2022
- Published by Taylor & Francis in North American Actuarial Journal
- Vol. 27 (4) , 675-688
Conformal Prediction Credibility Intervals
In the predictive modeling context, the credibility estimator is a point predictor; it is easy to calculate and avoids the model misspecification risk asymptotically, but it provides no quantification of inferential uncertainty. A Bayesian prediction interval quantifies uncertainty of prediction, but it often requires expensive computation and is subject to model misspecification risk even asymptotically. Is there a way to get the best of both worlds? Based on a powerful machine learning strategy called conformal prediction, this article proposes a method that converts the credibility estimator into a conformal prediction credibility interval. This conformal prediction credibility interval contains the credibility estimator, has computational simplicity, and guarantees finite-sample validity at a pre-assigned coverage level.
- 2 October 2022
- Published by Taylor & Francis in North American Actuarial Journal
- Vol. 27 (4) , 675-688
Conformal Prediction Credibility Intervals
In the predictive modeling context, the credibility estimator is a point predictor; it is easy to calculate and avoids the model misspecification risk asymptotically, but it provides no quantification of inferential uncertainty. A Bayesian prediction interval quantifies uncertainty of prediction, but it often requires expensive computation and is subject to model misspecification risk even asymptotically. Is there a way to get the best of both worlds? Based on a powerful machine learning strategy called conformal prediction, this article proposes a method that converts the credibility estimator into a conformal prediction credibility interval. This conformal prediction credibility interval contains the credibility estimator, has computational simplicity, and guarantees finite-sample validity at a pre-assigned coverage level.
- 16 December 2021
- Published by Taylor & Francis in North American Actuarial Journal
- Vol. 26 (4) , 485-495
Sample Size Determination for Credibility Estimation
The credibility estimator has been successfully and widely applied by practitioners in the insurance industry for decades. However, an important issue remains unresolved: What sample size should the actuary choose when he or she applies the credibility estimator? In particular, is the size of a given claims dataset large enough for the credibility estimator to be accurate? This article aims to address this issue by first suggesting a sample size criterion for the credibility estimator and then proposing a sample size based on that. The proposed sample size is easy to apply, and it guarantees provable accuracy for credibility estimation.
- 16 December 2021
- Published by Taylor & Francis in North American Actuarial Journal
- Vol. 26 (4) , 485-495
Sample Size Determination for Credibility Estimation
The credibility estimator has been successfully and widely applied by practitioners in the insurance industry for decades. However, an important issue remains unresolved: What sample size should the actuary choose when he or she applies the credibility estimator? In particular, is the size of a given claims dataset large enough for the credibility estimator to be accurate? This article aims to address this issue by first suggesting a sample size criterion for the credibility estimator and then proposing a sample size based on that. The proposed sample size is easy to apply, and it guarantees provable accuracy for credibility estimation.