In multi-armed bandit problems, information acquired from experimentation is valuable because it tells the agent whether to select a particular option again in the future. This article tests whether people undervalue this information because they are ambiguity averse, or have a distaste for uncertainty about the average quality of each alternative. It is shown that ambiguity averse agents have lower than optimal Gittins indexes, appearing to undervalue information from experimentation, but are willing to pay more than ambiguity neutral (...) agents to learn the true mean of the payoff distribution, appearing to overvalue objectively given information. This prediction is tested with a laboratory experiment that elicits a Gittins index and a willingness to pay on six two-armed bandits. Consistent with the predictions of ambiguity aversion, the Gittins indexes are significantly lower than optimal and willingnesses to pay are significantly higher than optimal. (shrink)

Decision making in noisy and changing environments requires a fine balance between exploiting knowledge about good courses of action and exploring the environment in order to improve upon this knowledge. We present an experiment on a restless bandit task in which participants made repeated choices between options for which the average rewards changed over time. Comparing a number of computational models of participants’ behavior in this task, we find evidence that a substantial number of them balanced exploration and exploitation (...) by considering the probability that an option offers the maximum reward out of all the available options. (shrink)

We consider a situation where an individual is facing an uncertain situation, but may costly alter his knowledge of the uncertainties. We study in this context how risk aversion may modify the individual search behavior. We consider a one-armed bandit problem (where one arm is safe and the other is risky) and study how the agent risk aversion can change the sequence of arms selected. The main result is that when the utility function is more concave, the agent has (...) more chances to select the safe arm. We also discuss how search is affected by risk aversion. (shrink)

We consider a situation where an individual is facing an uncertain situation, but may costly alter his knowledge of the uncertainties. We study in this context how risk aversion may modify the individual search behavior. We consider a one-armed bandit problem (where one arm is safe and the other is risky) and study how the agent risk aversion can change the sequence of arms selected. The main result is that when the utility function is more concave, the agent has (...) more chances to select the safe arm. We also discuss how search is affected by risk aversion. (shrink)

The aim of this paper is to study the monotonicity properties with respect to the probability distribution of the state processes, of optimal decisions in bandit decision problems. Orderings of dynamic discrete projects are provided by extending the notion of stochastic dominance to stochastic processes.

The aim of this paper is to study the monotonicity properties with respect to the probability distribution of the state processes, of optimal decisions in bandit decision problems. Orderings of dynamic discrete projects are provided by extending the notion of stochastic dominance to stochastic processes.

P. Clodius’ lex frumentaria of 58 BC tends to be viewed as a populist law that greatly exacerbated, if not largely created, the problems associated with the provision of grain in the 50s. Yet the law was intended to alleviate the high costs due to increased demand created by Cato’s lex frumentaria of 62. Moreover, it was not conceived as a one-shot, populist solution, but was part of a larger strategy aimed at mitigating the problem of supply. In ways (...) hitherto unrealised, Clodius’ annexation of Cyprus and the consular exemption of Delos from taxation formed part of a wider debate on tackling the root cause of high costs and unreliable supply: the resurgence of piracy in the eastern Mediterranean. (shrink)

In February 2020, as covid-19 infections spread to more than fifty countries, public health officials needed to recommend how the public could protect themselves, balancing safety and urgency. But there was very little data since this novel virus had only been identified three months prior. How could public health officials decide with insufficient data? The multi-armed bandit problem of computer science offers adaptive decision-making procedures that can achieve both safety and urgency. These adaptive methods balance learning information (exploring) with (...) using information (exploiting), adjusting the balance toward learning when uncertainty is high (March 1991; Kaelbling et al. 1996). Related methods are already used in adaptive clinical trials for pharmaceuticals (Pallmann et al. 2018). But we still need to develop these methods for non-pharmaceutical interventions, as I will illustrate with a case study of public mask-wearing to reduce the spread of covid-19. Public health pronouncements impact future learning. (shrink)

We study the wisdom of the crowd in three sequential decision‐making tasks: the Balloon Analogue Risk Task (BART), optimal stopping problems, and bandit problems. We consider a behavior‐based approach, using majority decisions to determine crowd behavior and show that this approach performs poorly in the BART and bandit tasks. The key problem is that the crowd becomes progressively more extreme as the decision sequence progresses, because the diversity of opinion that underlies the wisdom of the crowd (...) is lost. We also consider model‐based approaches to each task. This involves inferring cognitive models for each individual based on their observed behavior, and using these models to predict what each individual would do in any possible task situation. We show that this approach performs robustly well for all three tasks and has the additional advantage of being able to generalize to new problems for which there are no behavioral data. We discuss potential applications of the model‐based approach to real‐world sequential decision problems and discuss how our approach contributes to the understanding of collective intelligence. (shrink)

Multiarm bandit problems have been used to model the selection of competing scientific theories by boundedly rational agents. In this paper, I define a variable-arm bandit problem, which allows the set of scientific theories to vary over time. I show that Roth-Erev reinforcement learning, which solves multiarm bandit problems in the limit, cannot solve this problem in a reasonable time. However, social learning via preferential attachment combined with individual reinforcement learning which discounts the past, does.

We evaluate the asymptotic performance of boundedly-rational strategies in multi-armed bandit problems, where performance is measured in terms of the tendency (in the limit) to play optimal actions in either (i) isolation or (ii) networks of other learners. We show that, for many strategies commonly employed in economics, psychology, and machine learning, performance in isolation and performance in networks are essentially unrelated. Our results suggest that the appropriateness of various, common boundedly-rational strategies depends crucially upon the social context (...) (if any) in which such strategies are to be employed. (shrink)

Probability updating via Bayes' rule often entails extensive informational and computational requirements. In consequence, relatively few practical applications of Bayesian adaptive control techniques have been attempted. This paper discusses an alternative approach to adaptive control, Bayesian in spirit, which shifts attention from the updating of probability distributions via transitional probability assessments to the direct updating of the criterion function, itself, via transitional utility assessments. Results are illustrated in terms of an adaptive reinvestment two-armed bandit problem.

The laser chaos decision maker has been demonstrated to enable ultra-high-speed solutions of multiarmed bandit problems or decision-making in the GHz order. However, the underlying mechanisms are not well understood. In this paper, we analyze the chaotic dynamics inherent in experimentally observed laser chaos time series via surrogate data and further accelerate the decision-making performance via parameter optimization. We first evaluate the negative autocorrelation in a chaotic time series and its impact on decision-making detail. Then, we analyze the (...) decision-making ability using three different surrogate chaos time series to examine the underlying mechanism. We clarify that the negative autocorrelation of laser chaos improves decision-making and that the amplitude distribution of the original laser chaos time series is not optimal. Hence, we introduce a new parameter for adjusting the amplitude distribution of the laser chaos to enhance the decision-making performance. This study provides a new insight into exploiting the supremacy of chaotic dynamics in artificially constructed intelligent systems. (shrink)

The bounded confidence model has become a popular tool for studying communities of epistemically interacting agents. The model makes the idealizing assumption that all agents always have access to all other agents’ belief states. We draw on resources from network epistemology to do away with this assumption. In the model to be proposed, we impose an explicit communication network on a community, due to which each agent has access to the beliefs of only a selection of other agents. A much-discussed (...) result from network epistemology shows that densely connected communication networks are not always preferable to sparser networks. The aim of this paper is to investigate whether there are any noteworthy network effects in a version of the bounded confidence model augmented with communication networks, and in particular whether the aforementioned result from network epistemology can be replicated in that version. (shrink)

Through numbers of studies on the formation of equivalence relations and causal induction, it is known that human beings tend to consider conditional statements ``if p then q " as biconditional statements ``if and only if p then q ": we call the tendency to perceive ``if p then q " as ``if q then p " the ``symmetry bias". On the other hand, many studies on children's word learning have pointed out that children tend to expect each object has (...) only one label. This is so-called the ``mutual exclusivity bias". This bias implies. (shrink)

Ample manpower and material were mobilized for research on Liuhsia Chih in 1974. A thorough investigation of the life of a very famous leader of the working people's uprising and a clarification of erroneous historical legends about him are important for studies of history. Undoubtedly, a correct understanding of this historical character's features and propaganda about his revolutionary deeds will be conducive to the cause of socialism in our country. It is regrettable that some widely publicized works in 1974 were (...) rather inferior in quality and had unfavorable effects because their authors accepted the guidance of erroneous political thought and because the academic atmosphere was somewhat foul. As a result, people became confused. All this was caused by the control, interference and destruction of the "gang of four.". (shrink)

The aim of this paper is to clarify under what conditions a philosophical problem arises. I will describe two ways in which we might perceive a question as a problem. First, when we fnd ourselves inclined to believe in propositions that appear incompatible with each other. Second, when we fnd ourselves inclined to believe in propositions that seem incompatible with our desires. I will discuss both of these cases and articulate a didactic strategy – the Clash Strategy – which can (...) be used in order to plan and facilitate dialogic inquiry in the classroom. This strategy suggests introducing stimuli that will help students appreciate the clash that drives the philosophical question to be explored during the dialogue. In summary, this paper aims to ofer insights to all teachers interested in promoting dialogic inquiry, whether philosophical or otherwise, in education. (shrink)

Two types of formal models - landscape search tasks and two-armed bandit models - are often used to study the effects that various social factors have on epistemic performance. I argue that they can be understood within a single framework. In this unified framework, I develop a model that may be used to understand the effects of functional and demographic diversity and their interaction. Using the unified model, I find that the benefit of demographic diversity is most pronounced in (...) a functionally homogeneous group, and decreases with the increase of functional diversity. (shrink)

How does a government arise from anarchy? In a classic article, Mancur Olson theorized that it could occur when a roving bandit decides to settle down. This stationary bandit comes to recognize an encompassing interest in its territory, improving its lot by providing governing and committing to stable rates of theft. The bandits highlighted by Olson are not individuals but rather groups organized to act collectively. I provide a club-theoretic analysis of bandits. I characterize the violence as a (...) club good, i.e., one that is non-rival but from which out-group members can be excluded. I elaborate on the changes in the club’s group interest and in-group that will likely accompany the roving-to-stationary transition. To illustrate the salient points of the analysis, I provide a case study of the Visigoths: their emergence as a roving confederacy in the late fourth century, their migration and subsequent settlement, and the establishment of the Visigothic Kingdom. (shrink)

Here I consider the two most venerated arguments about the existence of God: the Ontological Argument and the Argument from Evil. The Ontological Argument purports to show that God’s nature guarantees that God exists. The Argument from Evil purports to show that God’s nature, combined with some plausible facts about the way the world is, guarantees (or is very compelling grounds for thinking) that God does not exist. Obviously, both arguments cannot be sound. But I argue here that they are (...) both unsound for the very same reason. (shrink)

As Vetter says, we are at the “beginning of the debate, not the end” (2015: 300) when it comes to evaluating her potentiality-based account of metaphysical modality. This paper contributes to this developing debate by highlighting three problems for Vetter’s account. Specifically, I begin (§1) by articulating some relevant details of Vetter’s potentiality-based view. This leads to the first issue (§2), concerning unclarity in the idea of degrees of potentiality. Similarly, the second issue (§3) raises trouble for Vetter’s proposed (...) individuation conditions for potentialities. Finally, the third issue (§4) is about apparently unmanifestable intrinsic potentialities, and suggests that there might be some deeper problems with anchoring metaphysical possibilities in concrete objects. More generally, though the issues detailed here are problematic, I do not take them to be fatal. However, they do show that, at minimum, further clarification of Vetter’s potentiality view is required. As Vetter says, we are at the “beginning of the debate, not the end” (2015: 300) when it comes to evaluating her potentiality-based account of metaphysical modality. This paper contributes to this developing debate by highlighting three problems for Vetter’s account. Specifically, I begin (§1) by articulating some relevant details of Vetter’s potentiality-based view. This leads to the first issue (§2), concerning unclarity in the idea of degrees of potentiality. Similarly, the second issue (§3) raises trouble for Vetter’s proposed individuation conditions for potentialities. Finally, the third issue (§4) is about apparently unmanifestable intrinsic potentialities, and suggests that there might be some deeper problems with anchoring metaphysical possibilities in concrete objects. More generally, though the issues detailed here are problematic, I do not take them to be fatal. However, they do show that, at minimum, further clarification of Vetter’s potentiality view is required. (shrink)

This book is an intercultural exploration of the full scope of evil. The problems of evil have beset humanity throughout the ages and continue to trouble us. The studies here examine evil in Asian thought, in Western theory, in the cosmic order, in human psychology, and in social practice. Insights are added to the philosophical discussions from religion, culture, history, law, technology, and literature.

There are arguments for determinism. Admittedly, this is opposed by the fact of everyday experience of autonomy. In the following, it is argued for the compatibility of determinism and autonomy. Taking up considerations of Donald MacKay, a fatalistic attitude can be refuted as false. Repeatedly, attempts have been made to defend the possibility of autonomy with reference to quantum physical indeterminacy. But its statistical randomness clearly misses the meaning of autonomy. What is decisive, on the other hand, is the possibility (...) of knowledge, which opens up opportunities for planning, freedom of choice and ultimately 'self-choice'. Results of neurobiological research, especially Benjamin Libet's and more recently John-Dylan Haynes', seem to refute this: Actions are unconsciously initiated before conscious decision. But, as Libet has also shown, consciousness always has the possibility of a veto – and thus also of knowledge-driven action control. Ultimately, the idea of possible self-choice can thus become the determining condition. Only such a form of rational self-determination establishes a spiritual identity and at the same time represents the maximum of autonomy possible for human beings. (shrink)

What corresponds to the present-day ‘transcendental-pragmatic’ concept of ultimate grounding in Hegel is his claim to absoluteness of the logic. Hegel’s fundamental intuition is that of a ‘backward going grounding’ obtaining the initially unproved presuppositions, thereby ‘wrapping itself into a circle’ – the project of the self-grounding of logic, understood as the self-explication of logic by logical means. Yet this is not about one of the multiple ‘logics’ which as formal constructs cannot claim absoluteness. It is rather a fundamental logic (...) that only makes logical textures possible at all and so owns transcendental character. The non-contradiction-principle is an example for this. Es- sential is that it is ‘under-cover-effcient’ as soon as meaningful concepts are used. Self-explication of the fundamental logic then means explicating its implicit under-cover validity, in fact by means of the fundamental logic itself. As is shown this is the affair of dialectic which thereby is to be understood as ultimate grounding of the fundamental logic. This is analyzed in detail using the example of the being/non-being-dialectic. As is demonstrated each explication step generates a new implicit issue and therewith a new explication-discrepancy inducing an antinomical structure that anew forwards the explication procedure. So this is entirely determined by itself. Decisive for the ultimate grounding argumentation is that thereby an objectively verifyable procedure is found, which is apparently possible only in a Hegelian framework. In contrast the immediate evidence of a speech act claimed by the transcendental-pragmatic position has only private character, which is grounding-theoretically irrelevant. (shrink)