Model-based Decision Research Articles

Computational and behavioral markers of model-based decision making in childhood.

Human decision-making is underpinned by distinct systems that differ in flexibility and associated cognitive cost. A widely accepted dichotomy distinguishes between a cheap but rigid model-free system and a flexible but costly model-based system. Typically, humans use a hybrid of both types of decision-making depending on environmental demands. However, children's use of a model-based system during decision-making has not yet been shown. While prior developmental work has identified simple building blocks of model-based reasoning in young children (1-4 years old), there has been little evidence of this complex cognitive system influencing behavior before adolescence. Here, by using a modified task to make engagement in cognitively costly strategies more rewarding, we show that children aged 5-11-years (N=85), including the youngest children, displayed multiple indicators of model-based decision making, and that the degree of its use increased throughout childhood. Unlike adults (N=24), however, children did not display adaptive arbitration between model-free and model-based decision-making. Our results demonstrate that throughout childhood, children can engage in highly sophisticated and costly decision-making strategies. However, the flexible arbitration between decision-making strategies might be a critically late-developing component in human development.

Abstract 5176: Tumor dynamic model-based decision support for phase Ib immunotherapy combination studies

Abstract Background: Tumor growth inhibition (TGI)-overall survival (OS) models predict OS distributions and study outcomes (OS hazard ratio (HR)) in a number of cancers and settings (Bruno, Clin Cancer Res 2020;26:1787-95). Model-derived TGI metrics such as tumor growth rate (KG) have been considered as a promising early endpoint compared to classical response criteria (Gong J Clin Oncol 2020; doi: 10.1200/JCO.2020.38.15_suppl.9541). This work is assessing the validity of tumor growth rate estimates to inform GO/noGO decisions for Phase Ib studies. Methods: We resampled baseline characteristics and longitudinal sum of longest diameters (SLD, RECIST 1.1) from a positive Phase III study (IMpower150, Socinski, N Engl J Med 2018;378:2288-301) to mimic a Phase Ib study of 15, 30, or 40 patients per arm (atezolizumab +bevacizumab+carboplatin+paclitaxel (A+BCP) vs. BCP) with 10 months recruitment and 6 or 3 months follow-up (FU) after the last patient recruited. Effect sizes across A+BCP vs. BCP or BCP vs. BCP were calculated as geometric mean ratio (GMR) for KG, difference (d) in objective response rate (ORR), and HR for progression-free survival (PFS) and simulated OS. For this, individual KG were estimated using a bi-exponential model (Claret, Clin Cancer Res 2018;24:3292-8); ORR and PFS were estimated per RECIST 1.1, OS was simulated using a previously published TGI-OS model (Chan CPT:PSP 2021; doi.org/10.1002/psp4.12686) with 400 patients per arm bootstrapped from the 15, 30 or 40 patients of the resampled subsamples. Correct and incorrect Go decisions were based on the probability (prob) to achieve target effect sizes in A+BCP vs. BCP and BCP vs. BCP, respectively, across 500 replicated subsamples. Results: To not exceed an incorrect Go decision of 20% and for 40 patients/6 months FU, correct Go based on prob(KG GMR<0.90), prob(dORR>0.10), prob(OS HR<0.85) and prob(PFS HR<0.70) were 83%, 69%, 67% and 58%, respectively. For 40 patients/3 months FU, 30 patients/6 or 3 months FU, the ranking did not change and prob(KG GMR<0.90) correct Go decision rate remained around 80%. Conclusions: This evaluation suggests that model-based estimates of on-treatment tumor growth rate could be used as an exploratory endpoint to inform early clinical decisions. Expansion of this work is planned in other settings e.g. single-arm studies commonly used in early oncology clinical development. Citation Format: René Bruno, Mathilde Marchand, Kenta Yoshida, Phyllis Chan, Haocheng Li, Wei Zou, Francois Mercier, Pascal Chanu, Benjamin Wu, Anthony Lee, Chunze Li, Jin Y. Jin. Tumor dynamic model-based decision support for phase Ib immunotherapy combination studies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 5176.

Need for cognition does not account for individual differences in metacontrol of decision making

Humans show metacontrol of decision making, that is they adapt their reliance on decision-making strategies toward situational differences such as differences in reward magnitude. Specifically, when higher rewards are at stake, individuals increase reliance on a more accurate but cognitively effortful strategy. We investigated whether the personality trait Need for Cognition (NFC) explains individual differences in metacontrol. Based on findings of cognitive effort expenditure in executive functions, we expected more metacontrol in individuals low in NFC. In two independent studies, metacontrol was assessed by means of a decision-making task that dissociates different reinforcement-learning strategies and in which reward magnitude was manipulated across trials. In contrast to our expectations, NFC did not account for individual differences in metacontrol of decision making. In fact, a Bayesian analysis provided moderate to strong evidence against a relationship between NFC and metacontrol. Beyond this, there was no consistent evidence for relationship between NFC and overall model-based decision making. These findings show that the effect of rewards on the engagement of effortful decision-making strategies is largely independent of the intrinsic motivation for engaging in cognitively effortful tasks and suggest a differential role of NFC for the regulation of cognitive effort in decision making and executive functions.

Model-Based Decision Support System for Improving Emergency Response

A Mobile-Based Ontology Integrated Decision Support System was built in an effort to improve the transparency in disaster situations. Using hypothetical chemical hazard scenarios, the decision-making accuracy and time taken by incident commanders using the Incident Command System was compared to incident commanders without the use of the system. The existing design of incident command systems limits the ability to manage workload and results in a limited span of knowledge. Sixteen firefighters were tested using the mobile-based ontology integrated decision support system. The results discussed in this study indicate that the information presented by the decision support system significantly improved commander’s decision-making. Potential applications of this research include Strengthening disaster risk reduction and management capacity through improved preparedness; Advancing regional and national initiatives for disaster response and risk reduction; and providing the foundation for future policies and procedures that can be easily implemented within the ICS framework.

Alcohol dependence decreases functional activation of the caudate nucleus during model-based decision processes.

Impaired decision making, a key characteristic of alcohol dependence (AD), manifests in continuous alcohol consumption despite severe negative consequences. The neural basis of this impairment in individuals with AD and differences with known neural decision mechanisms among healthy subjects are not fully understood. In particular, it is unclear whether the choice behavior among individuals with AD is based on a general impairment of decision mechanisms or is mainly explained by altered value attribution, with an overly high subjective value attributed to alcohol-related stimuli. Here, we use a functional magnetic resonance imaging (fMRI) monetary reward task to compare the neural processes of model-based decision making and value computation between AD individuals (n=32) and healthy controls (n=32). During fMRI, participants evaluated monetary offers with respect to dynamically changing constraints and different levels of uncertainty. Individuals with AD showed lower activation associated with model-based decision processes in the caudate nucleus than controls, but there were no group differences in value-related neural activity or task performance. Our findings highlight the role of the caudate nucleus in impaired model-based decisions of alcohol-dependent individuals.

The need for and conceptual design of an AI model-based Integrated Flight Advisory System

Automation and an increasing level of autonomy are not novelties in aircraft cockpits. One of the main goals of automation in aviation is to increase pilots’ situation awareness (SA) and reduce their workload – both of high importance for a safe flight. Increasing automation has historically reduced aviation accident rates and improved efficiency. Yet, currently implemented systems have also contributed to accidents, when failed, or were insufficient to increase pilots’ SA for improving their decision-making. This paper discusses the need for an enhanced decision-support system and its potential benefits for safer aviation. A model-based decision support system that leverages artificial intelligence, Integrated Flight Advisory System (IFAS), is presented. Further, the conceptual design of this system is described. The overview and roadmap provided in this paper consist of an effort toward a more holistic, pilot-centered, AI-based decision-support that can contribute to safer aviation.

Cost-effectiveness of artificial intelligence screening for diabetic retinopathy in rural China

BackgroundDiabetic retinopathy (DR) has become a leading cause of global blindness as a microvascular complication of diabetes. Regular screening of diabetic retinopathy is strongly recommended for people with diabetes so that timely treatment can be provided to reduce the incidence of visual impairment. However, DR screening is not well carried out due to lack of eye care facilities, especially in the rural areas of China. Artificial intelligence (AI) based DR screening has emerged as a novel strategy and show promising diagnostic performance in sensitivity and specificity, relieving the pressure of the shortage of facilities and ophthalmologists because of its quick and accurate diagnosis. In this study, we estimated the cost-effectiveness of AI screening for DR in rural China based on Markov model, providing evidence for extending use of AI screening for DR.MethodsWe estimated the cost-effectiveness of AI screening and compared it with ophthalmologist screening in which fundus images are evaluated by ophthalmologists. We developed a Markov model-based hybrid decision tree to analyze the costs, effectiveness and incremental cost-effectiveness ratio (ICER) of AI screening strategies relative to no screening strategies and ophthalmologist screening strategies (dominated) over 35 years (mean life expectancy of diabetes patients in rural China). The analysis was conducted from the health system perspective (included direct medical costs) and societal perspective (included medical and nonmedical costs). Effectiveness was analyzed with quality-adjusted life years (QALYs). The robustness of results was estimated by performing one-way sensitivity analysis and probabilistic analysis.ResultsFrom the health system perspective, AI screening and ophthalmologist screening had incremental costs of $180.19 and $215.05 but more quality-adjusted life years (QALYs) compared with no screening. AI screening had an ICER of $1,107.63. From the societal perspective which considers all direct and indirect costs, AI screening had an ICER of $10,347.12 compared with no screening, below the cost-effective threshold (1–3 times per capita GDP of Chinese in 2019).ConclusionsOur analysis demonstrates that AI-based screening is more cost-effective compared with conventional ophthalmologist screening and holds great promise to be an alternative approach for DR screening in the rural area of China.

Model-based learning retrospectively updates model-free values

Reinforcement learning (RL) is widely regarded as divisible into two distinct computational strategies. Model-free learning is a simple RL process in which a value is associated with actions, whereas model-based learning relies on the formation of internal models of the environment to maximise reward. Recently, theoretical and animal work has suggested that such models might be used to train model-free behaviour, reducing the burden of costly forward planning. Here we devised a way to probe this possibility in human behaviour. We adapted a two-stage decision task and found evidence that model-based processes at the time of learning can alter model-free valuation in healthy individuals. We asked people to rate subjective value of an irrelevant feature that was seen at the time a model-based decision would have been made. These irrelevant feature value ratings were updated by rewards, but in a way that accounted for whether the selected action retrospectively ought to have been taken. This model-based influence on model-free value ratings was best accounted for by a reward prediction error that was calculated relative to the decision path that would most likely have led to the reward. This effect occurred independently of attention and was not present when participants were not explicitly told about the structure of the environment. These findings suggest that current conceptions of model-based and model-free learning require updating in favour of a more integrated approach. Our task provides an empirical handle for further study of the dialogue between these two learning systems in the future.

A model-based decision framework for the multi-depot multi-travelling salesman problem with split and delivery demand considering different key performance indicators

A model-based decision framework for the multi-depot multi-travelling salesman problem with split and delivery demand considering different key performance indicators

Strategic Earning on Tokenized Platforms via Model-based Decision Making

Strategic Earning on Tokenized Platforms via Model-based Decision Making

Valence bias in metacontrol of decision making in adolescents and young adults.

The development of metacontrol of decision making and its susceptibility to framing effects were investigated in a sample of 201 adolescents and adults in Germany (12-25years, 111 female, ethnicity not recorded). In a task that dissociates model-free and model-based decision making, outcome magnitude and outcome valence were manipulated. Both adolescents and adults showed metacontrol and metacontrol tended to increase across adolescence. Furthermore, model-based decision making was more pronounced for loss compared to gain frames but there was no evidence that this framing effect differed with age. Thus, the strategic adaptation of decision making continues to develop into young adulthood and for both adolescents and adults, losses increase the motivation to invest cognitive resources into an effortful decision-making strategy.

Model blindness: A Framework for Understanding how Model-Based Decision Support Systems can Lead to Performance Degradation

Model-Based Decision Support Systems (MDSS) are ubiquitous in many high-consequence domains like emergency management and military command and control. This paper proposes a framework of model-blindness imposed by MDSS via filtering relevant information and presenting irrelevant information to the decision-maker, causing performance degradation. The paper also defines novel categories of performance: model-limited, strategy-limited, and context-limited. We propose a theoretical confluence model of operator performance under model blindness and a framework to evaluate model blindness empirically. We report simulation results to demonstrate the impact of model blindness on performance and outline mitigation strategies for designers and developers of MDSS. The recommendations and insights can be used to evaluate the limitations associated with particular MDSS before implementation.

High Performance Classification Model to Identify Ransomware Payments for Heterogeneous Bitcoin Networks

The Bitcoin cryptocurrency is a worldwide prevalent virtualized digital currency conceptualized in 2008 as a distributed transactions system. Bitcoin transactions make use of peer-to-peer network nodes without a third-party intermediary, and the transactions can be verified by the node. Although Bitcoin networks have exhibited high efficiency in the financial transaction systems, their payment transactions are vulnerable to several ransomware attacks. For that reason, investigators have been working on developing ransomware payment identification techniques for bitcoin transactions’ networks to prevent such harmful cyberattacks. In this paper, we propose a high performance Bitcoin transaction predictive system that investigates the Bitcoin payment transactions to learn data patterns that can recognize and classify ransomware payments for heterogeneous bitcoin networks. Specifically, our system makes use of two supervised machine learning methods to learn the distinguishing patterns in Bitcoin payment transactions, namely, shallow neural networks (SNN) and optimizable decision trees (ODT). To validate the effectiveness of our solution approach, we evaluate our machine learning based predictive models on a recent Bitcoin transactions dataset in terms of classification accuracy as a key performance indicator and other key evaluation metrics such as the confusion matrix, positive predictive value, true positive rate, and the corresponding prediction errors. As a result, our superlative experimental result was registered to the model-based decision trees scoring 99.9% and 99.4% classification detection (two-class classifier) and accuracy (multiclass classifier), respectively. Hence, the obtained model accuracy results are superior as they surpassed many state-of-the-art models developed to identify ransomware payments in bitcoin transactions.

Mathematical model of the feedback between global supply chain disruption and COVID-19 dynamics

The pandemic of COVID-19 has become one of the greatest threats to human health, causing severe disruptions in the global supply chain, and compromising health care delivery worldwide. Although government authorities sought to contain the spread of SARS-CoV-2, by restricting travel and in-person activities, failure to deploy time-sensitive strategies in ramping-up of critical resource production exacerbated the outbreak. Here, we developed a mathematical model to analyze the effects of the interaction between supply chain disruption and infectious disease dynamics using coupled production and disease networks built on global data. Analysis of the supply chain model suggests that time-sensitive containment strategies could be created to balance objectives in pandemic control and economic losses, leading to a spatiotemporal separation of infection peaks that alleviates the societal impact of the disease. A lean resource allocation strategy can reduce the impact of supply chain shortages from 11.91 to 1.11% in North America. Our model highlights the importance of cross-sectoral coordination and region-wise collaboration to optimally contain a pandemic and provides a framework that could advance the containment and model-based decision making for future pandemics.

A Model-Based Tool for Assessing the Impact of Land Use Change Scenarios on Flood Risk in Small-Scale River Systems—Part 1: Pre-Processing of Scenario Based Flood Characteristics for the Current State of Land Use

Land use changes influence the water balance and often increase surface runoff. The resulting impacts on river flow, water level, and flood should be identified beforehand in the phase of spatial planning. In two consecutive papers, we develop a model-based decision support system for quantifying the hydrological and stream hydraulic impacts of land use changes. Part 1 presents the semi-automatic set-up of physically based hydrological and hydraulic models on the basis of geodata analysis for the current state. Appropriate hydrological model parameters for ungauged catchments are derived by a transfer from a calibrated model. In the regarded lowland river basins, parameters of surface and groundwater inflow turned out to be particularly important. While the calibration delivers very good to good model results for flow (Evol =2.4%, R = 0.84, NSE = 0.84), the model performance is good to satisfactory (Evol = −9.6%, R = 0.88, NSE = 0.59) in a different river system parametrized with the transfer procedure. After transferring the concept to a larger area with various small rivers, the current state is analyzed by running simulations based on statistical rainfall scenarios. Results include watercourse section-specific capacities and excess volumes in case of flooding. The developed approach can relatively quickly generate physically reliable and spatially high-resolution results. Part 2 builds on the data generated in part 1 and presents the subsequent approach to assess hydrologic/hydrodynamic impacts of potential land use changes.

Development of an Intelligent Decision Support System for Attaining Sustainable Growth within a Life Insurance Company

Consumer behaviour is one of the most important and complex areas of research. It acknowledges the buying behaviour of consumer clusters towards any product, such as life insurance policies. Among various factors, the three most well-known determinants on which human conjecture depends for preferring a product are demographic, economic and psychographic factors, which can help in developing an accurate market design and strategy for the sustainable growth of a company. In this paper, the study of customer satisfaction with regard to a life insurance company is presented, which focused on comparing artificial intelligence-based, data-driven approaches to classical market segmentation approaches. In this work, an artificial intelligence-based decision support system was developed which utilises the aforementioned factors for the accurate classification of potential buyers. The novelty of this paper lies in developing supervised machine learning models that have a tendency to accurately identify the cluster of potential buyers with the help of demographic, economic and psychographic factors. By considering a combination of the factors that are related to the demographic, economic and psychographic elements, the proposed support vector machine model and logistic regression model-based decision support systems were able to identify the cluster of potential buyers with collective accuracies of 98.82% and 89.20%, respectively. The substantial accuracy of a support vector machine model would be helpful for a life insurance company which needs a decision support system for targeting potential customers and sustaining its share within the market.

A framework for parameter estimation using sharp-interface seawater intrusion models

Sharp-interface seawater intrusion models present shorter run times than variable density codes, which makes them practical for regional, decision-support groundwater modeling. Although parameter estimation and uncertainty analyses are essential steps for model-based decision making, their implementation in seawater intrusion models has remained limited. Few guidelines are available and it is unclear which observations should be used, what processing they require and what weighting strategy should be used. A framework has been developed for parameter estimation using a regional sharp-interface decision-support model applied to a real-world example in the Magdalen Islands (Quebec, Canada). This framework included the assimilation of head observations collected from shallow wells, deep open wells and pumping wells, as well as freshwater-seawater interface observations derived from deep open wells, TDEM (time-domain electromagnetic) and ERT (electrical resistivity tomography) geophysical surveys. A model was developed using MODFLOW-SWI2 in which fast model run times allowed the estimation of numerous parameters, including a heterogeneous hydraulic conductivity field with pilot points. Following parameter estimation with PEST, the uncertainty of several model forecasts, i.e. the volume of freshwater and the interface elevation near municipal pumping wells, was examined with the first-order second moment (FOSM) approach and a data worth analysis was carried out. While the observations presented a low signal-to-noise ratio, parameter estimation was effective to reduce the uncertainty of model forecasts. Interface observations, and particularly geophysical observations, were most useful to reduce predictive uncertainties. Head and interface observations from deep open wells were biased and could not be suitably reproduced by the model. The framework developed here is relatively straightforward and could be implemented more systematically. The study also provides recommendations to guide future data collection strategies in coastal aquifers.

A simulation model-based clinical decision tool to guide personalized treatment based on individual characteristics: Does 21-gene recurrence score assay testing change decisions?

e12507 Background: There is a need for web-based decision tools that integrate clinicopathologic features and genomic information to guide breast cancer therapy for women with node-negative, hormone receptor positive, HER2 negative (“early-stage”) breast cancer. We developed a novel simulation model-based clinical decision tool that provides prognostic estimates of treatment outcomes based on age, tumor size, grade, and comorbidities with and without 21-gene recurrence scores (RS). Methods: We adapted an extant breast cancer simulation model developed within the NCI-funded Cancer Intervention and Surveillance Modeling Network (CISNET) to derive estimates for the 10-year risks of distant recurrence, breast cancer-specific mortality, other cause mortality and life-years gained with endocrine vs. chemo-endocrine therapy for individual women based on their age, tumor size, grade, and comorbidity-level with and without RS test results. The model used an empiric Bayesian analytical approach to combine information from clinical trials, registry and claims data to provide individual estimates. External validation of the model was performed by comparing model-based breast cancer mortality rates and observed rates in the Surveillance Epidemiology and End Results (SEER) registry. Results: Several exemplar profiles were selected to illustrate the clinical utility of the decision tool. For example, the absolute chemotherapy benefit for 10-year distant recurrence risk and life-years gained, without RS testing, and the outcomes if a woman got tested and had a RS 16-20 are provided below for a 40-44-year-old woman and a 65–69-year-old woman diagnosed with a small (≤2cm), intermediate grade tumor and mild comorbidities. Conclusions: Simulation modeling is useful for creating clinical decision tools to support shared decision making for early-stage breast cancer treatment.[Table: see text]

Experimental testing of a model-based decision support system integrated with smart sensors to optimize irrigation strategies for processing tomato: a case study in southern Italy

Experimental testing of a model-based decision support system integrated with smart sensors to optimize irrigation strategies for processing tomato: a case study in southern Italy

Coordinated Strategy for a Model-Based Decision Support Tool for Coronavirus Disease, Utah, USA.

The coronavirus disease pandemic has highlighted the key role epidemiologic models play in supporting public health decision-making. In particular, these models provide estimates of outbreak potential when data are scarce and decision-making is critical and urgent. We document the integrated modeling response used in the US state of Utah early in the coronavirus disease pandemic, which brought together a diverse set of technical experts and public health and healthcare officials and led to an evidence-based response to the pandemic. We describe how we adapted a standard epidemiologic model; harmonized the outputs across modeling groups; and maintained a constant dialogue with policymakers at multiple levels of government to produce timely, evidence-based, and coordinated public health recommendations and interventions during the first wave of the pandemic. This framework continues to support the state’s response to ongoing outbreaks and can be applied in other settings to address unique public health challenges.