Generalisations of a Bayesian decision-theoretic randomisation procedure and the impact of delayed responses

Williamson, S. Faye; Jacko, Peter; Jaki, Thomas

doi:10.1016/j.csda.2021.107407

Cited by 9 publications

(3 citation statements)

References 73 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the efficiency test of the decision method, this paper takes the time-consuming decision process (index 1) as the measurement standard, tests and evaluates the efficiency of the serial version of the adaptive decision method and the parallel search adaptive decision method, the composite adaptive decision method [24], the rule-based adaptive decision method [25] and the utility-based adaptive method [26] in the damage and replacement scenario. e parallel version refers to using parallel optimization mechanism to execute this method in a multi-threaded manner by dividing the population.…”

Section: Comparative Experimental Testsmentioning

confidence: 99%

Adaptive Decision Method in C3I System

Gao

Wang

Nazarko

et al. 2022

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

As an important system in the national defense and military information construction, the command, control, communication, and intelligence (C3I) system urgently needs to establish an adaptive process to deal with the dynamic operating environment and changeable task requirements to ensure the long-term effective and stable operation of the system. As an important part of this process, the adaptive decision method should have the ability of online trade-off decision. Therefore, this paper presents an adaptive decision method based on parallel computing and optimization theory. This method combines operational requirements and commander preference to achieve the parallel adaptive decision solution. The experimental results show that the presented decision method can generate online trade-off strategies to deal with typical command and control scenarios of damage replacement in a simulated environment, effectively guide the system to carry out adjustment behavior, and achieve the goal of dynamic response to environmental changes and task changes.

show abstract

Section: Comparative Experimental Testsmentioning

confidence: 99%

Adaptive Decision Method in C3I System

Gao

Wang

Nazarko

et al. 2022

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

show abstract

“…It was shown that the corresponding performance could be improved by imputation based on a short-term predictor. Williamson et al [64] identify the existing gap between the theory and clinical practice for Bayesian response-adaptive procedures by considering whether responses are intermediately available or delayed, respectively. However, these findings may not apply to the missing data problem in the context of responseadaptive designs, since missing values can be viewed as a very extreme form of delay where the outcomes would never be available [43].…”

Section: Introductionmentioning

confidence: 99%

Some performance considerations when using multi-armed bandit algorithms in the presence of missing data

Chen,

Lee,

Villar

et al. 2022

Preprint

View full text Add to dashboard Cite

When using multi-armed bandit algorithms, the potential impact of missing data is often overlooked. In practice, the simplest approach is to ignore missing outcomes and continue to sample following the bandit algorithm. We investigate the impact of missing data on several bandit algorithms via a simulation study assuming the rewards are missing at random. We focus on two-armed bandit algorithms with binary outcomes in the context of patient allocation for clinical trials with relatively small sample sizes. However, our results can apply to other applications of bandit algorithms where missing data is expected to occur. We assess the resulting operating characteristics, including the expected reward (i.e., allocation results). Different probabilities of missingness in both arms are considered. The key finding of our work is that when using the simplest strategy of ignoring missing data, the corresponding impact on the performance of multi-armed bandit strategies varies according to their way of balancing the exploration-exploitation trade-off. Algorithms that are geared towards exploration continue to assign samples to the arm with more missing responses, and this arm is perceived as the superior arm by the algorithm. By contrast, algorithms that are geared towards exploitation would do the opposite and not assign samples to the arms with more missing responses. Furthermore, for algorithms focusing more on exploration, we illustrate that the problem of missing responses can be alleviated using a simple mean imputation approach.

show abstract

“…It was shown that the corresponding performance could be improved by imputation based on a short-term predictor. Williamson et al [ 29 ] identify the existing gap between the theory and clinical practice for Bayesian response-adaptive procedures by considering whether responses are intermediately available or delayed, respectively. However, these findings may not apply to the missing data problem in the context of response-adaptive designs, since missing values can be viewed as a very extreme form of delay where the outcomes would never be available [ 30 ].…”

Section: Introductionmentioning

confidence: 99%

Some performance considerations when using multi-armed bandit algorithms in the presence of missing data

et al. 2022

View full text Add to dashboard Cite

When comparing the performance of multi-armed bandit algorithms, the potential impact of missing data is often overlooked. In practice, it also affects their implementation where the simplest approach to overcome this is to continue to sample according to the original bandit algorithm, ignoring missing outcomes. We investigate the impact on performance of this approach to deal with missing data for several bandit algorithms through an extensive simulation study assuming the rewards are missing at random. We focus on two-armed bandit algorithms with binary outcomes in the context of patient allocation for clinical trials with relatively small sample sizes. However, our results apply to other applications of bandit algorithms where missing data is expected to occur. We assess the resulting operating characteristics, including the expected reward. Different probabilities of missingness in both arms are considered. The key finding of our work is that when using the simplest strategy of ignoring missing data, the impact on the expected performance of multi-armed bandit strategies varies according to the way these strategies balance the exploration-exploitation trade-off. Algorithms that are geared towards exploration continue to assign samples to the arm with more missing responses (which being perceived as the arm with less observed information is deemed more appealing by the algorithm than it would otherwise be). In contrast, algorithms that are geared towards exploitation would rapidly assign a high value to samples from the arms with a current high mean irrespective of the level observations per arm. Furthermore, for algorithms focusing more on exploration, we illustrate that the problem of missing responses can be alleviated using a simple mean imputation approach.

show abstract

Generalisations of a Bayesian decision-theoretic randomisation procedure and the impact of delayed responses

Cited by 9 publications

References 73 publications

Adaptive Decision Method in C3I System

Adaptive Decision Method in C3I System

Some performance considerations when using multi-armed bandit algorithms in the presence of missing data

Some performance considerations when using multi-armed bandit algorithms in the presence of missing data

Contact Info

Product

Resources

About