Adversarial vulnerabilities of human decision-making

Dezfouli, Amir; Nock, Richard; Dayan, Peter

doi:10.1073/pnas.2016921117

Cited by 21 publications

(29 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In cognitive neuroscience, another approach to this problem is to use behavioural experimental tools to explain the model’s behaviour [6,10]. One way to carry out this task is by examining the different experimental settings that make the model fail, known as adversarial examples, [24], which has a long tradition in cognitive psychology, from the use of visual illusions to study perception to the characterisation of biases in decision making [43]. Another method is to train many models with different goals, and to examine which models best describe human behaviour [23].…”

Section: Discussionmentioning

confidence: 99%

“…Several researchers used different approaches to overcome this problem. One approach was to train many different models with different goals, and examine how they perform in predicting human behaviour, thus controlling for the model’s goal [23], and another approach was to use adversarial examples that meant misleading a model and thus gaining insights on its operations [24]. We suggest another direction, which is to use tools from cognitive neuroscience, the same explicit cognitive models described above, to characterise and explain the operations of such a data-driven black box model.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Using Deep Learning to Predict Human Decisions, and Cognitive Models to Explain Deep Learning Models

Fintz

Osadchy

Hertz

2021

Preprint

View full text Add to dashboard Cite

Deep neural networks (DNN) models have the potential to provide new insights in the study of human decision making, due to their high capacity and data-driven design. While these models may be able to go beyond theory-driven models in predicting human behaviour, their opaque nature limits their ability to explain how an operation is carried out. This explainability problem remains unresolved. Here we demonstrate the use of a DNN model as an exploratory tool to identify predictable and consistent human behaviour in value-based decision making beyond the scope of theory-driven models. We then propose using theory-driven models to characterise the operation of the DNN model. We trained a DNN model to predict human decisions in a four-armed bandit task. We found that this model was more accurate than a reinforcement-learning reward-oriented model geared towards choosing the most rewarding option. This disparity in accuracy was more pronounced during times when the expected reward from all options was similar, i.e., no unambiguous good option. To investigate this disparity, we introduced a reward-oblivious model, which was trained to predict human decisions without information about the rewards obtained from each option. This model captured decision-sequence patterns made by participants (e.g., a-b-c-d). In a series of experimental offline simulations of all models we found that the general model was in line with a reward-oriented model’s predictions when one option was clearly better than the others.However, when options’ expected rewards were similar to each other, it was in-line with the reward-oblivious model’s pattern completion predictions. These results indicate the contribution of predictable but task-irrelevant decision patterns to human decisions, especially when task-relevant choices are not immediately apparent. Importantly, we demonstrate how theory-driven cognitive models can be used to characterise the operation of DNNs, making them a useful explanatory tool in scientific investigation.Author SummaryDeep neural networks (DNN) models are an extremely useful tool across multiple domains, and specifically for performing tasks that mimic and predict human behaviour. However, due to their opaque nature and high level of complexity, their ability to explain human behaviour is limited. Here we used DNN models to uncover hitherto overlooked aspects of human decision making, i.e., their reliance on predictable patterns for exploration. For this purpose, we trained a DNN model to predict human choices in a decision-making task. We then characterised this data-driven model using explicit, theory-driven cognitive models, in a set of offline experimental simulations. This relationship between explicit and data-driven approaches, where high-capacity models are used to explore beyond the scope of established models and theory-driven models are used to explain and characterise these new grounds, make DNN models a powerful scientific tool.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Using Deep Learning to Predict Human Decisions, and Cognitive Models to Explain Deep Learning Models

Fintz

Osadchy

Hertz

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…This privacy issue can be further extended to the sport business space Dezfouli et al ( 2020 ) have shown how AI can be designed to manipulate human behaviour. Algorithms learned from humans' responses who were participating in controlled experiments.…”

Section: The Future Of Ai In Sportmentioning

confidence: 99%

Artificial Intelligence and Machine Learning in Sport Research: An Introduction for Non-data Scientists

Chmait

Westerbeek

2021

Front. Sports Act. Living

View full text Add to dashboard Cite

In the last two decades, artificial intelligence (AI) has transformed the way in which we consume and analyse sports. The role of AI in improving decision-making and forecasting in sports, amongst many other advantages, is rapidly expanding and gaining more attention in both the academic sector and the industry. Nonetheless, for many sports audiences, professionals and policy makers, who are not particularly au courant or experts in AI, the connexion between artificial intelligence and sports remains fuzzy. Likewise, for many, the motivations for adopting a machine learning (ML) paradigm in sports analytics are still either faint or unclear. In this perspective paper, we present a high-level, non-technical, overview of the machine learning paradigm that motivates its potential for enhancing sports (performance and business) analytics. We provide a summary of some relevant research literature on the areas in which artificial intelligence and machine learning have been applied to the sports industry and in sport research. Finally, we present some hypothetical scenarios of how AI and ML could shape the future of sports.

show abstract

“…The RL agent's goal is to intervene in its environment of human users to learn an optimal policy maximizing the accumulation of designer-specified rewards (e.g., clicks). Although platforms have only recently begun to apply RL to interact with and control digital environments of human users, researchers have demonstrated empirically that RL agents can induce humans to exhibit a variety of "target behaviors" selected for by algorithm designers 64 . A growing literature now details how RL-based systems can fail-often due to poorly specified reward functions-and how these failures may have social consequences 65 .…”

Section: Academic Data Science Research In the Realm Of Behavior Modificationmentioning

confidence: 99%

Barriers for Academic Data Science Research in the New Realm of Behavior Modification by Digital Platforms

2021

View full text Add to dashboard Cite

The era of behavioral big data has created new avenues for data science research, with many new contributions stemming from academic researchers. Yet, data controlled by platforms has become increasingly difficult for academic researchers to access. Platforms now routinely use behavior modification techniques to manipulate users' behavior, leaving academic researchers further isolated in conducting important data science and computational social science research. This isolation results from researchers' lack of access to human behavioral data, and crucially, to both the data on machine behavior that triggers and learns from the human data and the platform's behavior modification mechanisms. Given the impact of behavior modification on individual and societal well-being, we discuss the consequences for scientific knowledge creation and the roles and responsibilities of academic data scientists. Finally, we consider how current and future regulatory actions may enable academic data scientists to conduct meaningful research on platform environments in the realm of behavior modification.

show abstract

Adversarial vulnerabilities of human decision-making

Cited by 21 publications

References 14 publications

Using Deep Learning to Predict Human Decisions, and Cognitive Models to Explain Deep Learning Models

Using Deep Learning to Predict Human Decisions, and Cognitive Models to Explain Deep Learning Models

Artificial Intelligence and Machine Learning in Sport Research: An Introduction for Non-data Scientists

Barriers for Academic Data Science Research in the New Realm of Behavior Modification by Digital Platforms

Contact Info

Product

Resources

About