Error is protean, ubiquitous and crucial in scientific process. In this paper it is argued that understanding scientific process requires what is currently absent: an adaptable, context-sensitive functional role for error in science that naturally harnesses error identification and avoidance to positive, success-driven, science. This paper develops a new account of scientific process of this sort, error and success driving Self-Directed Anticipative Learning (SDAL) cycling, using a recent re-analysis of ape-language research as test example. The example shows the limitations (...) of other accounts of error, in particular Mayo’s (Error and the growth of experimental knowledge, 1996) error-statistical approach, and SDAL cycling shows how they can be fruitfully contextualised. (shrink)

Sequence learning is fundamental to a wide range of cognitive functions. Explaining how sequences—and the relations between the elements they comprise—are learned is a fundamental challenge to cognitive science. However, although hundreds of articles addressing this question are published each year, the actual learning mechanisms involved in the learning of sequences are rarely investigated. We present three experiments that seek to examine these mechanisms during a typing task. Experiments 1 and 2 tested learning during typing single (...) letters on each trial. Experiment 3 tested for “chunking” of these letters into “words.” The results of these experiments were used to examine the mechanisms that could best account for them, with a focus on two particular proposals: statistical transitional probability learning and discriminative error-driven learning. Experiments 1 and 2 showed that error-driven learning was a better predictor of response latencies than either n-gram frequencies or transitional probabilities. No evidence for chunking was found in Experiment 3, probably due to interspersing visual cues with the motor response. In addition, learning occurred across a greater distance in Experiment 1 than Experiment 2, suggesting that the greater predictability that comes with increased structure leads to greater learnability. These results shed new light on the mechanism responsible for sequence learning. Despite the widely held assumption that transitional probability learning is essential to this process, the present results suggest instead that the sequences are learned through a process of discriminative learning, involving prediction and feedback from prediction error. (shrink)

Linguistic category learning has been shown to be highly sensitive to linear order, and depending on the task, differentially sensitive to the information provided by preceding category markers (premarkers, e.g., gendered articles) or succeeding category markers (postmarkers, e.g., gendered suffixes). Given that numerous systems for marking grammatical categories exist in natural languages, it follows that a better understanding of these findings can shed light on the factors underlying this diversity. In two discriminative learning simulations and an artificial language (...) learning experiment, we identify two factors that modulate linear order effects in linguistic category learning: category structure and the level of abstraction in a category hierarchy. Regarding category structure, we find that postmarking brings an advantage for learning category diagnostic stimulus dimensions, an effect not present when categories are non‐confusable. Regarding levels of abstraction, we find that premarking of super‐ordinate categories (e.g., noun class) facilitates learning of subordinate categories (e.g., nouns). We present detailed simulations using a plausible candidate mechanism for the observed effects, along with a comprehensive analysis of linear order effects within an expectation‐based account of learning. Our findings indicate that linguistic category learning is differentially guided by pre‐ and postmarking, and that the influence of each is modulated by the specific characteristics of a given category system. (shrink)

Linguistic category learning has been shown to be highly sensitive to linear order, and depending on the task, differentially sensitive to the information provided by preceding category markers (premarkers, e.g., gendered articles) or succeeding category markers (postmarkers, e.g., gendered suffixes). Given that numerous systems for marking grammatical categories exist in natural languages, it follows that a better understanding of these findings can shed light on the factors underlying this diversity. In two discriminative learning simulations and an artificial language (...) learning experiment, we identify two factors that modulate linear order effects in linguistic category learning: category structure and the level of abstraction in a category hierarchy. Regarding category structure, we find that postmarking brings an advantage for learning category diagnostic stimulus dimensions, an effect not present when categories are non‐confusable. Regarding levels of abstraction, we find that premarking of super‐ordinate categories (e.g., noun class) facilitates learning of subordinate categories (e.g., nouns). We present detailed simulations using a plausible candidate mechanism for the observed effects, along with a comprehensive analysis of linear order effects within an expectation‐based account of learning. Our findings indicate that linguistic category learning is differentially guided by pre‐ and postmarking, and that the influence of each is modulated by the specific characteristics of a given category system. (shrink)

A novel data-driven model-free adaptive control approach is first proposed by combining the advantages of model-free adaptive control and data-driven optimal iterative learning control, and then its stability and convergence analysis is given to prove algorithm stability and asymptotical convergence of tracking error. Besides, the parameters of presented approach are adaptively adjusted with fuzzy logic to determine the occupied proportions of MFAC and DDOILC according to their different control performances in different control stages. Lastly, the proposed fuzzy DDMFAC approach (...) is applied to the control of particle quality in drug development phase of spray fluidized-bed granulation process, and its control effect is compared with MFAC and DDOILC and their fuzzy forms, in which the parameters of MFAC and DDOILC are adaptively adjusted with fuzzy logic. The effectiveness of the presented FDDMFAC approach is verified by a series of simulations. (shrink)

Unsupervised learning algorithms are widely used for many important statistical tasks with numerous applications in science and industry. Yet despite their prevalence, they have attracted remarkably little philosophical scrutiny to date. This stands in stark contrast to supervised and reinforcement learning algorithms, which have been widely studied and critically evaluated, often with an emphasis on ethical concerns. In this article, I analyze three canonical unsupervised learning problems: clustering, abstraction, and generative modeling. I argue that these methods raise (...) unique epistemological and ontological questions, providing data-driven tools for discovering natural kinds and distinguishing essence from contingency. This analysis goes some way toward filling the lacuna in contemporary philosophical discourse on unsupervised learning, as well as bringing conceptual unity to a heterogeneous field more often described by what it isnot(i.e., supervised or reinforcement learning) than by what itis. I submit that unsupervised learning is not just a legitimate subject of philosophical inquiry but perhaps the most fundamental branch of all AI. However, an uncritical overreliance on unsupervised methods poses major epistemic and ethical risks. I conclude by advocating for a pragmatic, error-statistical approach that embraces the opportunities and mitigates the challenges posed by this powerful class of algorithms. (shrink)

Human motor learning is governed by a suite of interacting mechanisms each one of which modifies behavior in distinct ways and rely on different neural circuits. In recent years, much attention has been given to one type of motor learning, called motor adaptation. Here, the field has generally focused on the interactions of three mechanisms: sensory prediction error SPE-driven, explicit, and reinforcement learning. Studies of these mechanisms have largely treated them as modular, aiming to model how the (...) outputs of each are combined in the production of overt behavior. However, when examined closely the results of some studies also suggest the existence of additional interactions between the sub-components of each learning mechanism. In this perspective, we propose that these sub-component interactions represent a critical means through which different motor learning mechanisms are combined to produce movement; understanding such interactions is critical to advancing our knowledge of how humans learn new behaviors. We review current literature studying interactions between SPE-driven, explicit, and reinforcement mechanisms of motor learning. We then present evidence of sub-component interactions between SPE-driven and reinforcement learning as well as between SPE-driven and explicit learning from studies of people with cerebellar degeneration. Finally, we discuss the implications of interactions between learning mechanism sub-components for future research in human motor learning. (shrink)

Artificial intelligence, as a nonhuman entity, is increasingly used to inform, direct, or supplant nursing care and clinical decision‐making. The boundaries between human‐ and nonhuman‐driven nursing care are blurred with the advent of sensors, wearables, camera devices, and humanoid robots at such an accelerated pace that the critical evaluation of its influence on patient safety has not been fully assessed. Since the pivotal release of To Err is Human, patient safety is being challenged by the dynamic healthcare environment like never (...) before, with nursing at a critical juncture to steer the course of artificial intelligence integration in clinical decision‐making. This paper presents an overview of artificial intelligence and its application in healthcare and highlights the implications which affect nursing as a profession, including perspectives on nursing education and training recommendations. The legal and policy challenges which emerge when artificial intelligence influences the risk of clinical errors and safety issues are discussed. (shrink)

This study first investigates robust iterative learning control issue for a class of two-dimensional linear discrete singular Fornasini–Marchesini systems under iteration-varying boundary states. Initially, using the singular value decomposition theory, an equivalent dynamical decomposition form of 2-D LDSFM is derived. A simple P-type ILC law is proposed such that the ILC tracking error can be driven into a residual range, the bound of which is relevant to the bound parameters of boundary states. Specially, while the boundary states of 2-D (...) LDSFM satisfy iteration-invariant boundary states, accurate tracking on 2-D desired surface trajectory can be accomplished by using 2-D linear inequality theory. In addition, extension to 2-D LDSFM without direct transmission from inputs to outputs is presented. A numerical example is used to illustrate the effectiveness and feasibility of the designed ILC law. (shrink)

Context: Public universities in South Africa are currently facing the challenge of decolonising knowledge. This change requires a review of curriculums, as well as teaching and learning with the goal of embracing the epistemology of the learners, addressing issues such as social justice and transformation. Problem: Human communication is subject to several perceptual errors in both listening and seeing, which challenges the success of the communication in the education system. The ability of the teacher and the learners to effectively (...) communicate with one another is a factor for the success of each reaching their goals. The teacher imparts her knowledge in the classroom, but according to von Foerster, “[i]t is the listener, not the speaker, who determines the meaning of an utterance,” for the listener contextualises this information based on her own past lived experience. Thus, the student’s epistemology and her expression of her understanding is integral in the classroom context and should be actively included into the education system. Method: I present a cybernetic approach to the teacher-learner system, challenging traditional ideas about the role of each actor within the system, with special attention given to Pask’s conversation theory. Results: Early empirical findings suggest that a conversational contextual approach results in higher student involvement and better memory retention among the learners. Conversational approaches that are epistemologically inclusive diffuse social problems where the student groups require their individual worldviews to be reflected within the curriculum. This reduces the friction of competing epistemologies within the education system, moving toward a co-created contextually-driven knowledge system. Implications: Many educators would like deeper engagement from their learners but have not found a way to successfully engage the student group. A cybernetic approach is one method that can be adopted to remedy this. This is particularly useful in contexts where there is cultural diversity and impending social change. Constructivist content: I address von Glasersfeld’s points on human cognition, linking it to Austin’s speech acts. (shrink)

In order to model any macroscopic system, it is necessary to aggregate both spatially and taxonomically. If average processes are assumed, then kinetic equations of “population dynamics” can be derived. Much effort has gone into showing the important effects introduced by non-average effects (fluctuations) in generating symmetry-breaking transitions and creating structure and form. However, the effects of microscopic diversity have been largely neglected. We show that evolution will select for populations which retain “variability,” even though this is, at any given (...) time, loss-making, predicting that we shall not observe populations with “optimal behavior,” but populations which can “learn.” This lesser short-term efficiency may be why natural diversity is so great. Evolution is seen to be “driven” by the noise to which it leads. (shrink)

Given the current popularization of computer programming and the trends of informatization and digitization, colleges have actively responded by making programming lessons compulsory for students of all disciplines. However, students from different ethnic groups often have different learning responses to such lessons due to their respective cultural backgrounds, the environment in which they grew up, and their consideration for future employment. In this study, an AI-assisted programming module was developed and used to compare the differences between multi-ethnic college students (...) in terms of their theoretical and actual learning expectancy, motivation, and effectiveness. The module conducted analysis through the deep learning network and examined the relevant processes that the students underwent during programming lessons, as well as the types of errors they had committed. Their learning motivation for and actual learning performance in programming were then examined based on the cognitive learning theory. The results of the experiment, which involved 96 multi-ethnic college students, indicated that the two groups had dissimilar theoretical performance in terms of their expectancy and motivation for learning programming. The indigenous students’ main concern was whether programming would affect their families or tribes, and this concern affected and was reflected in their learning outcomes. In contrast, the learning motivation and goals of Han Chinese students were driven by the cognition of the value of programming to themselves. The research findings can contribute toward the cognition and understanding of multi-ethnic students when learning computer programming and development of the appropriate teaching methods, and serve as a reference for subsequent research on integrating multiculturalism into computer programming lessons. (shrink)

Dans le cadre de ce travail, nous présentons une expérimentation menée auprès d’un public allophone inscrit à une formation universitaire. Ce travail a pour objectif de relever, d’une part, les spécificités dans les productions orales de ce public, et d’autre part, de démontrer l’intérêt d’un apprentissage sur corpus afin de construire un exposé structuré. Cette étude permet de s’ouvrir à d’autres perspectives didactiques en partant d’un corpus d’apprenants.

Syntactic priming in language production is the increased likelihood of using a recently encountered syntactic structure. In this paper, we examine two theories of why speakers can be primed: error‐driven learning accounts (Bock, Dell, Chang, & Onishi, 2007; Chang, Dell, & Bock, 2006) and activation‐based accounts (Pickering & Branigan, 1999; Reitter, Keller, & Moore, 2011). Both theories predict that speakers should be primed by the syntactic choices of others, but only activation‐based accounts predict that speakers should be able (...) to prime themselves. Here we test whether speakers can be primed by their own productions in three behavioral experiments and find evidence of structural persistence following both comprehension and speakers’ own productions. We also find that comprehension‐based priming effects are larger for rarer syntactic structures than for more common ones, which is most consistent with error‐driven accounts. Because neither error‐driven accounts nor activation‐based accounts fully explain the data, we propose a hybrid model. (shrink)

Research shows that exploring language corpora through data-driven learning plays a significant role in language learning. Nevertheless, it is not clear if using concordancing as an application of DDL affects the learners’ second language motivation. To address this gap, the current study adopted a triangulation design, validating quantitative data model, and a quasi-experimental design. Ninety English-major university students with an intermediate level of English language proficiency, divided into control and experimental groups, took part in the study. Drawing on (...) a second language motivational self-system questionnaire, the findings of the study did not result in any statistically significant differences between the students in their second language motivational self-system. Altogether, the students found learning English through a DDL approach with concordancing less motivating than receiving explicit instruction. The study has implications for language teaching and learning. (shrink)

Rogers & McClelland (R&M) present a powerful account of semantic (conceptual) learning. Their model admirably handles many characteristics of early concept formation, but it also needs to address attentional biases, and distinguish direct input from error-driven learning, and fast versus slow learning. Not distinguishing implicit and explicit knowledge means that the authors also cannot explain why some coherently varying information becomes accessible and other information does not.

In this chapter, we explore three social functions of emotion, which parallel three interpretations of Herman Melville's Bartleby. We argue that emotions can serve as commitment devices, which nudge individuals toward social conformity and thereby increase the likelihood of ongoing cooperation. We argue that emotions can play a role in Machiavellian strategies, which help us get away with norm violations. And we argue that emotions can motivate social recalibration, by alerting us to systemic social failures. In the second half of (...) the chapter, we then argue that emotions guide behavior by attuning us to our social environment through a process of error-driven learning. We suggest that their most basic social function is to reveal tensions between individuals and social norms; and we contend that because emotions are socially scaffolded, they leave open multiple strategies for resolving such tensions. Specifically, we argue that differences in conceptualization, as well as differences in circumstance, can generate emotions that nudge us toward cooperation, lead us to adopt Machiavellian strategies, or motivate us to engage in forms of social recalibration. Bartleby’s despair signaled a mismatch between his expectations and the expectations of Wall Street, and each of these three strategies could have moderated his misery. (shrink)

Curiosity motivates the search for missing information, driving learning, scientific discovery, and innovation. Yet, identifying that there is a gap in one's knowledge is itself a critical step, and may demand that one formulate a question to precisely express what is missing. Our work captures the integral role of self‐generated questions during the acquisition of new information, which we refer to as active‐curiosity‐driven learning. We tested active‐curiosity‐driven learning using our “Curiosity Question & Answer Task” paradigm, where participants (...) (N=135) were asked to generate questions in response to novel, incomplete factual statements and provided the opportunity to forage for answers. We also introduce new measures of question quality that express how well questions capture stimulus and foraging information. We hypothesized that active question asking should influence behavior across the stages of our task by increasing the probability that participants express curiosity, forage for answers, and remember what they had thereby discovered. We found that individuals who asked a high number of quality questions experienced elevated curiosity, were more likely to pursue missing information that was semantically related to their questions, and more likely to retain the information on a later cued recall test. Additional analyses revealed that curiosity played a predominant role in motivating participants to forage for missing information, and that both curiosity and satisfaction with the acquired information boosted memory recall. Overall, our results suggest that asking questions enhances the value of missing information, with important implications for learning and discovery of all forms. (shrink)

Everyone agrees that testing hypotheses is important, but Bowers et al. provide scant details about where hypotheses about perception and brain function should come from. We suggest that the answer lies in considering how information about the outside world could be acquired – that is, learned – over the course of evolution and development. Deep neural networks (DNNs) provide one tool to address this question.

We analyze naturally occurring datasets from student use of educational technologies to explore a long-standing question of the scope of transfer of learning. We contrast a faculty theory of broad transfer with a component theory of more constrained transfer. To test these theories, we develop statistical models of them. These models use latent variables to represent mental functions that are changed while learning to cause a reduction in error rates for new tasks. Strong versions of these models provide (...) a common explanation for the variance in task difficulty and transfer. Weak versions decouple difficulty and transfer explanations by describing task difficulty with parameters for each unique task. We evaluate these models in terms of both their prediction accuracy on held-out data and their power in explaining task difficulty and learning transfer. In comparisons across eight datasets, we find that the component models provide both better predictions and better explanations than the faculty models. Weak model variations tend to improve generalization across students, but hurt generalization across items and make a sacrifice to explanatory power. More generally, the approach could be used to identify malleable components of cognitive functions, such as spatial reasoning or executive functions. (shrink)

Purpose. The purpose of this study is to investigate how professionals learn from varying experiences with errors in health-care digitalization and develop and use negative knowledge and digital ignorance in efforts to improve digitalized health care. Design/methodology/approach. A two-year qualitative field study was conducted in the context of a public health-care organization working with digital patient communication. The data consisted of participant observation, semistructured interviews and document data. Inductive coding and a theoretically informed generation of themes were applied. Findings. The (...) findings show that both health-care and digital communication professionals learn through experiences with digital “rule-” and “knowledge-based” errors in patient communication and develop negative knowledge and awareness of digital ignorance. In their joint efforts, they use negative knowledge to “bend the rules” and to explore digital ignorance in efforts to improve patient communication. Originality/value. This study provides insight into the importance of collaboration between professionals with varying experience of errors in digitalizing patient communication. Such collaboration is required to acknowledge own shortcomings and create complementary negative knowledge to improve digital patient communication. This is particularly important when working with innovative digitalization in health care. (shrink)

A group of engineering students at the Technical University of Berlin, Germany, designed a course on engineering ethics. The core element of the developed Blue Engineering course are self-contained teaching-units, “building blocks”. These building blocks typically cover one complex topic and make use of various teaching methods using moderators who lead discussions, rather than experts who lecture. Consequently, the students themselves started to offer the credited course to their fellow students who take an active role in further developing the course (...) themselves. (shrink)

This article argues that errors in audio data processing should be examined to explore and expose the underlying components that enable linguistic communication and cross-cultural understanding. Examples of errors in the transcription of a Mexican social network’s conversations are analyzed to demonstrate the potential of such data in the development of sociocognitive language-processing theories. It is suggested that researchers working with audio-recorded data should expand the scope of what is considered useful data for the sake of both methodological reflexivity and (...) identifying the underlying cognitive processes enabling linguistic understanding. (shrink)

In their target article, McKay & Dennett (M&D) conclude that only “positive illusions” are adaptive misbeliefs. Relying on overly strict conceptual schisms (deficit vs. motivational, functional vs. organic, perception vs. belief), they prematurely discount delusions asbiologicallyadaptive. In contrast to their view that “motivation” plays a psychological but not a biological function in a two-factor model of the forming and maintenance of delusions, we propose asingleimpairment in prediction-error–driven (i.e., motivational) learning in three stages in which delusions play a biologically adaptive (...) role. (shrink)

The error statistical account of testing uses statistical considerations, not to provide a measure of probability of hypotheses, but to model patterns of irregularity that are useful for controlling, distinguishing, and learning from errors. The aim of this paper is (1) to explain the main points of contrast between the error statistical and the subjective Bayesian approach and (2) to elucidate the key errors that underlie the central objection raised by Colin Howson at our PSA 96 Symposium.

Position papers on artificial intelligence (AI) ethics are often framed as attempts to work out technical and regulatory strategies for attaining what is commonly called trustworthy AI. In such papers, the technical and regulatory strategies are frequently analyzed in detail, but the concept of trustworthy AI is not. As a result, it remains unclear. This paper lays out a variety of possible interpretations of the concept and concludes that none of them is appropriate. The central problem is that, by framing (...) the ethics of AI in terms of trustworthiness, we reinforce unjustified anthropocentric assumptions that stand in the way of clear analysis. Furthermore, even if we insist on a purely epistemic interpretation of the concept, according to which trustworthiness just means measurable reliability, it turns out that the analysis will, nevertheless, suffer from a subtle form of anthropocentrism. The paper goes on to develop the concept of strange error, which serves both to sharpen the initial diagnosis of the inadequacy of trustworthy AI and to articulate the novel epistemological situation created by the use of AI. The paper concludes with a discussion of how strange error puts pressure on standard practices of assessing moral culpability, particularly in the context of medicine. (shrink)

In recent years, self-paced learning has attracted much attention due to its improvement to nonconvex optimization based machine learning algorithms. As a methodology introduced from human learning, SPL dynamically evaluates the learning difficulty of each sample and provides the weighted learning model against the negative effects from hard-learning samples. In this study, we proposed a cognitive driven SPL method, i.e., retrospective robust self-paced learning, which is inspired by the following two issues in human (...) learning process: the misclassified samples are more impressive in upcoming learning, and the model of the follow-up learning process based on large number of samples can be used to reduce the risk of poor generalization in initial learning phase. We simultaneously estimated the degrees of learning-difficulty and misclassified in each step of SPL and proposed a framework to construct multilevel SPL for improving the robustness of the initial learning phase of SPL. The proposed method can be viewed as a multilayer model and the output of the previous layer can guide constructing robust initialization model of the next layer. The experimental results show that the R2SPL outperforms the conventional self-paced learning models in classification task. (shrink)

Research on human causal learning has largely focused on strength learning, or on computational-level theories; there are few formal algorithmic models of how people learn causal structure from covariations. We introduce a model that learns causal structure in a local manner via prediction-error learning. This local learning is then integrated dynamically into a unified representation of causal structure. The model uses computationally plausible approximations of rational learning, and so represents a hybrid between the associationist and (...) rational paradigms in causal learning research. We conclude by showing that the model provides a good fit to data from a previous experiment. (shrink)

This study used case studies from three science teachers to compare three groups of students studying Grade 8 physics using Robolab instead of traditional lab materials. The three teachers represented an English as a second language class, a regular class with many English language learner students, and a Mathematics, Engineering, Science Achievement class of afterschool volunteer students. The teachers responded to nine questions regarding issues such as how robotics addresses the middle school physics standards, promotes inquiry learning and science (...) literacy in students, increases applications and connection of science to everyday life, and provides the scaffolding for more discussion of everyday English and science academic language through the use of robotics materials and lab groups. The results indicated that robotics can promote inquiry, make physics more interesting, and help students learn and practice new words in English, among others. (shrink)

The correct use of connectives has great influence on language learners’ writing proficiency, while errors of connectives are common in foreign learners’ interlanguages. This study examines the types of errors that occur in native English-speaking learners’ Chinese writing, the possible causes for the errors, and the learners’ consequent learning strategies. The present research adopted corpora investigation, questionnaire survey, and focus-group interviews to examine the error types, causes of identified errors, and related learning strategies. Data analysis indicated that: the (...) main error types made by native English-speaking learners from high to low are misuse, overuse, mismatch, misplacement, and underuse of connectives; causes related to intralingual transfer greatly contributes to the presence of errors; and memory, social, and cognitive strategies were the most preferred, followed by metacognitive and compensation strategies, and then by effective strategies which were the least preferred. These findings showed that different strategies can be employed to cope with different errors in writing. The study further suggests that teachers and educators need to help native English-speaking learners find strategies that work best for them in terms of learning Chinese connectives. (shrink)