Dissociable learning processes,associative theory,and testimonial reviews: A comment on Smith and Church (2018)

Psychonomic Bulletin & Review - Smith and Church (Psychonomic Bulletin & Review, 25, 1565–1584 2018) present a “testimonial” review of dissociable learning processes...     

A Bayesian-optimal principle for learner-friendly adaptation in learning games

Adaptive learning games should provide opportunities for the student to learn as well as motivate playing until goals have been reached. In this paper, we give a mathematically rigorous treatment of the problem in the framework of Bayesian decision theory. To quantify the opportunities for learning, we assume that the learning tasks that yield the most information about the current skills of the student, while being desirable for measurement in their own right, would also be among those that are efficient for learning. Indeed, optimization of the expected information gain appears to naturally avoid tasks that are exceedingly demanding or exceedingly easy as their results are predictable and thus uninformative. Still, tasks that are efficient for learning may be experienced as too challenging, and the resulting failures can lower motivation. Therefore, in addition to quantifying the expected informational benefit for learning of any prospective task to be presented next, we also model the expected motivational cost of its presentation, measured simply as the estimated probability of failure in our example model. We propose a “learner-friendly” adaptation algorithm that chooses the learning tasks by optimizing the expected benefit divided by the expected cost. We apply this algorithm to a Rasch-like student model implemented within a real-world application and present initial results of a pilot experiment.     

Markov processes in learning theory

Consideration is given mathematical problems arising in two learning theories—one developed by Bush and Mosteller, the other developed by Estes. The theory of Bush and Mosteller leads to a class of Markov processes which have been studied in considerable detail (see [1] and [7]). The Estes model can be treated as a Markov chain, i.e., a Markov process with a finite number of states. For an important class of special cases, it is shown that the Bush-Mosteller model is, in a sense, a limiting form of the Estes model. The limiting probability distributions are derived for the cases treated in both models.This research was supported by the National Science Foundation through a grant given to the Dartmouth Mathematics Project.     

Conditional discrimination learning: A critique and amplification

Carter and Werner recently reviewed the literature on conditional discrimination learning by pigeons, which consists of studies of matching-to-sample and oddity-from-sample. They also discussed three models of such learning: the “multiple-rule” model (learning of stimulus-specific relations), the “configuration” model, and the “single-rule” model (concept learning). Although their treatment of the multiple-rule model, which seems most applicable to the pigeon data, is generally excellent, their discussion of the other two models is incomplete and sometimes inaccurate. Potential problems of terminology are discussed in the present paper, as are additional lines of research that deserve consideration by those interested in further work in this area. The issue of response versus stimulus selection (configuration versus compound-cue learning) is discussed in connection with the configuration model. Particular attention is given to Carter and Werner's criticism of the application, in studies with other species, of the learning set procedure in testing for single-rule learning. Some of the important related issues are: the bias for improvement on new problems in a series, the adequacy of a multiple-rule model to explain learning set formation, and evidence in favor of the single-rule model, at least in primates. Consideration of these additional contributions to the study of conditional discrimination learning emphasizes the usefulness of this task in the comparative study of cognitive processes.     

Aging,working memory,and discrimination learning

Older adults easily learn probabilistic relationships between cues and outcomes when the predictive event is the occurrence of a cue, but have greater difficulty when the predictive event is the nonoccurrence of a cue (Mutter & Pliske, 1996; Mutter & Plumlee, 2004; Mutter & Williams, 2004). This study explored whether this age-related deficit occurs in a simpler learning context and whether it might be related to working memory (WM) decline. We gave younger and older adults simultaneous discrimination tasks that allowed us to compare their ability to learn deterministic relationships when either the occurrence (feature positive; FP) or the nonoccurrence (feature negative; FN) of a distinctive feature predicted reinforcement. We also included a group of younger adults who performed the discrimination tasks under a concurrent WM load. Both age and WM load had a detrimental effect on initial FP and FN discrimination; however, these effects persisted only in FN discrimination after additional learning experience. Learning predictive relationships requires inductive reasoning processes that apparently do not operate as efficiently in individuals with reduced WM capacity. The impact of WM decline may ultimately be greater for negative cue–outcome relationships because learning these relationships requires more difficult inductive reasoning processes, which place greater demands on WM.     

Analyzing behavior implied by EWA learning: An emphasis on distinguishing reinforcement from belief learning

An important issue in the field of learning is to what extent one can distinguish between behavior resulting from either belief or reinforcement learning. Previous research suggests that it is difficult or even impossible to distinguish belief from reinforcement learning: belief and reinforcement models often fit the empirical data equally well. However, previous research has been confined to specific games in specific settings. In the present study we derive predictions for behavior in games using the EWA learning model (e.g., Camerer & Ho, 1999), a model that includes belief learning and a specific type of reinforcement learning as special cases. We conclude that belief and reinforcement learning can be distinguished, even in 2×2 games. Maximum differentiation in behavior resulting from either belief or reinforcement learning is obtained in games with pure Nash equilibria with negative payoffs and at least one other strategy combination with only positive payoffs. Our results help researchers to identify games in which belief and reinforcement learning can be discerned easily.     

A case for restricted-domain relational learning

Monkeys and pigeons learned a same/different task with pairs that were selected from a training set of eight picture stimuli. They showed no novel-stimulus transfer and hence no abstract-concept learning. They were also tested with novel pairs of the eight training pictures (i.e., combinations that had not been used in training) and with inverted pictures of the training pairs. If the subjects had learned the task item-specifically (e.g., if—then or configural learning), they should have failed these tests, but they performed well with novel combinations of training pictures and inverted pictures, suggesting that they learned the task relationally (i.e., on the basis of the relationship between the two pictures that were presented in each trial). This somewhat paradoxical conclusion of relational learning in the absence of abstract-concept learning is contrary to most theories of abstract-concept learning. The implications of this conclusion are discussed in the context of restricted-domain relational learning.     

Optimal Hierarchical Learning Path Design With Reinforcement Learning

E-learning systems are capable of providing more adaptive and efficient learning experiences for learners than traditional classroom settings. A key component of such systems is the learning policy. The learning policy is an algorithm that designs the learning paths or rather it selects learning materials for learners based on information such as the learners’ current progresses and skills, learning material contents. In this article, the authors address the problem of finding the optimal learning policy. To this end, a model for learners’ hierarchical skills in the E-learning system is first developed. Based on the hierarchical skill model and the classical cognitive diagnosis model, a framework to model various mastery levels related to hierarchical skills is further developed. The optimal learning path in consideration of the hierarchical structure of skills is found by applying a model-free reinforcement learning method, which does not require any assumption about learners’ learning transition processes. The effectiveness of the proposed framework is demonstrated via simulation studies.     

Dissociable behavioural outcomes of visual statistical learning

ABSTRACT

Statistical learning refers to the extraction of probabilistic relationships between stimuli and is increasingly used as a method to understand learning processes. However, numerous cognitive processes are sensitive to the statistical relationships between stimuli and any one measure of learning may conflate these processes; to date little research has focused on differentiating these processes. To understand how multiple processes underlie statistical learning, here we compared, within the same study, operational measures of learning from different tasks that may be differentially sensitive to these processes. In Experiment 1, participants were visually exposed to temporal regularities embedded in a stream of shapes. Their task was to periodically detect whether a shape, whose contrast was staircased to a threshold level, was present or absent. Afterwards, they completed a search task, where statistically predictable shapes were found more quickly. We used the search task to label shape pairs as “learned” or “non-learned”, and then used these labels to analyse the detection task. We found a dissociation between learning on the search task and the detection task where only non-learned pairs showed learning effects in the detection task. This finding was replicated in further experiments with recognition memory (Experiment 2) and associative learning tasks (Experiment 3). Taken together, these findings are consistent with the view that statistical learning may comprise a family of processes that can produce dissociable effects on different aspects of behaviour.     

Deep learning and punctuated equilibrium theory

Deep learning is associated with the latest success stories in AI. In particular, deep neural networks are applied in increasingly different fields to model complex processes. Interestingly, the underlying algorithm of backpropagation was originally designed for political science models. The theoretical foundations of this approach are very similar to the concept of Punctuated Equilibrium Theory (PET). The article discusses the concept of deep learning and shows parallels to PET. A showcase model demonstrates how deep learning can be used to provide a missing link in the study of the policy process: the connection between attention in the political system (as inputs) and budget shifts (as outputs).     

A neural-network model enabling sensorimotor learning: Application to the control of arm movements and some implications for speech-motor control and stuttering

Summary Low-level motor control is defined as adapting an organism to the unique physical properties of its own limbs. The two-jointed arm serves to exemplify that effective low-level motor control demands a neurally mediated inversion of the dynamics, as well as of the kinematics, of a limb system. Reflex-like processing — that is, feedforword of either actual or predicted proprioceptive signals —is thereby assumed to be the principle of the dynamics control. As regards speech-motor control, the overall tool transformation is assumed to transform the force pattern of the articulatory muscles into speech sounds. Like the arm model, the vocal-tract transformation thus defined is also divided into two parts, namely the transformation relating the muscle forces to the mechanospatial states of the vocal tract (which is analogous to the forward dynamics including natural interarticulatory couplings), and the transformation relating the mechanospatial states to the speech sounds. Low-level speech-motor control, then, needs to invert both transformations, each of which can be learned by means of the self-imitation algorithm. Erroneous learning can fail to decouple interarticulatory coupling and therefore lead to abnormal feedback loops through the reflex-like operating neural network, which in turn can cause stuttering if audiophonatoric coupling is involved in learning.     

Dissociation of binding and learning processes

A single encounter of a stimulus together with a response can result in a short-lived association between the stimulus and the response [sometimes called an event file, see Hommel, Müsseler, Aschersleben, & Prinz, (2001) Behavioral and Brain Sciences, 24, 910–926]. The repetition of stimulus–response pairings typically results in longer lasting learning effects indicating stimulus–response associations (e.g., Logan & Etherton, (1994) Journal of Experimental Psychology: Learning, Memory, and Cognition, 20, 1022–1050]. An important question is whether or not what has been described as stimulus–response binding in action control research is actually identical with an early stage of incidental learning (e.g., binding might be seen as single-trial learning). Here, we present evidence that short-lived binding effects can be distinguished from learning of longer lasting stimulus–response associations. In two experiments, participants always responded to centrally presented target letters that were flanked by response irrelevant distractor letters. Experiment 1 varied whether distractors flanked targets on the horizontal or vertical axis. Binding effects were larger for a horizontal than for a vertical distractor-target configuration, while stimulus configuration did not influence incidental learning of longer lasting stimulus–response associations. In Experiment 2, the duration of the interval between response n – 1 and presentation of display n (500 ms vs. 2000 ms) had opposing influences on binding and learning effects. Both experiments indicate that modulating factors influence stimulus–response binding and incidental learning effects in different ways. We conclude that distinct underlying processes should be assumed for binding and incidental learning effects.

     

Abstraction and model evaluation in category learning

Thirty previously published data sets, from seminal category learning tasks, are reanalyzed using the varying abstraction model (VAM). Unlike a prototype-versus-exemplar analysis, which focuses on extreme levels of abstraction only, a VAM analysis also considers the possibility of partial abstraction. Whereas most data sets support no abstraction when only the extreme possibilities are considered, we show that evidence for abstraction can be provided using the broader view on abstraction provided by the VAM. The present results generalize earlier demonstrations of partial abstraction (Vanpaemel & Storms, 2008), in which only a small number of data sets was analyzed. Following the dominant modus operandi in category learning research, Vanpaemel and Storms evaluated the models on their best fit, a practice known to ignore the complexity of the models under consideration. In the present study, in contrast, model evaluation not only relies on the maximal likelihood, but also on the marginal likelihood, which is sensitive to model complexity. Finally, using a large recovery study, it is demonstrated that, across the 30 data sets, complexity differences between the models in the VAM family are small. This indicates that a (computationally challenging) complexity-sensitive model evaluation method is uncalled for, and that the use of a (computationally straightforward) complexity-insensitive model evaluation method is justified.     

A novel machine learning approach for early detection of hepatocellular carcinoma patients

Liver cancer is quite common type of cancer among individuals worldwide. Hepatocellular carcinoma (HCC) is the malignancy of liver cancer. It has high impact on individual’s life and investigating it early can decline the number of annual deaths. This study proposes a new machine learning approach to detect HCC using 165 patients. Ten well-known machine learning algorithms are employed. In the preprocessing step, the normalization approach is used. The genetic algorithm coupled with stratified 5-fold cross-validation method is applied twice, first for parameter optimization and then for feature selection. In this work, support vector machine (SVM) (type C-SVC) with new 2level genetic optimizer (genetic training) and feature selection yielded the highest accuracy and F₁-Score of 0.8849 and 0.8762 respectively. Our proposed model can be used to test the performance with huge database and aid the clinicians.     

The 28th Bartlett Memorial Lecture Causal learning: An associative analysis

The concordance between performance and judgements of the causal effectiveness of an instrumental action suggests that such actions are mediated by causal knowledge. Although causal learning exhibits many associative phenomena—blocking, inhibitory or preventative learning, and super-learning—judgements of the causal status of a cue can be changed retrospectively as a result of learning episodes that do not directly involve the cue. In order to explain retrospective revaluation, a modified associative theory is described in which the learning processes for retrieved cue representations are the opposite to those for presented cues, and this theory is evaluated by studies of the role of within-compound associations in retrospective revaluation and blocking. However, this modified theory only applies when the within-compound association represents a contiguous rather than a causal cue relationship.     

Human temporal behavior and discrimination-reversal learning

The present study demonstrates that a relationship exists between individual differences in temporal behavior and individual differences in human discrimination-reversal learning behavior. When the results of performance on a time-estimation task employing the method of reproduction are compared with the results of acquisition performance on a complex form of discrimination-reversal learning task is can be demonstrated that underestimation of time is associated with faster learning and overestimation of time is associated with slower learning of the discrimination task. The experimental design was based on the historical fact that the Sechenov-Pavlov and Spence-Hull formulations assigned a primary role in learning to excitation and inhibition as intervening variables between the input of stimulation and output of response. The present study also used the tasks of time estimation and discrimination learning as dependent variables; and the concept that underestimation of time is associated with and underlying predominance of excitatory processes as a hypothetical construct. The conclusion was reached that one of the many theoretical processes which may be used to explain discrimination learning is the concept that acquisition of the correct response may be viewed as a function of the individual rates at which excitatory processes come to be conditioned to a predominance over existing inhibitory processes.     

Optional shift,and discrimination learning with redundant relevant dimensions: Predictions of an attentional model

The Zeaman-House-Lovejoy (ZHL) model for discrimination learning involves interacting perceptual and response learning processes, with rate parameters ? and θ, respectively. This paper describes a number of predictions of the ZHL model for the optional shift paradigm. Kendler and Kendler showed that the proportion of subjects who learn an optional shift as a reversal increases with subject's age, and Campione and others have argued that this trend may reflect developmental variation in parameters like ? and θ. Our main result is a formula specifying the dependence of reversal probability on ? and θ for the ZHL model.     

Varenicline ameliorates ethanol-induced deficits in learning in C57BL/6 mice

Ethanol is a frequently abused drug that impairs cognitive processes such as learning. Varenicline, an α4β2 nicotinic receptor partial agonist and α7 nicotinic receptor full agonist prescribed for smoking cessation, has been shown to decrease ethanol consumption. The current study investigated whether varenicline could ameliorate ethanol-induced deficits in learning and whether varenicline alters blood alcohol concentration in C57BL/6 mice. Conditioning consisted of two auditory conditioned stimulus (CS; 30 s, 85 dB white noise)—foot shock unconditioned stimulus (US; 2 s, 0.57 mA) pairings. For all studies, saline or ethanol (1.0, 1.5, 2.0 g/kg i.p.) was administered 15 min before training, and saline or varenicline (0.05, 0.1, 0.2 mg/kg i.p.) was administered 60 min before either training or testing. For blood alcohol analysis, saline or varenicline (0.1 mg/kg) was administered 60 min before collection, and saline or ethanol (1.0, 1.5, 2.0 g/kg) was administered 15 min before collection. Varenicline dose-dependently ameliorated ethanol-induced conditioning deficits for all three doses of ethanol when administered before training but not when administered 24 h later, before testing. In addition, varenicline did not alter blood alcohol concentration. The smoking cessation aid varenicline may have therapeutic uses for treating ethanol-associated disruptions in cognitive processes.     

Implicit learning in rule induction and problem solving

Using the Chinese Ring Puzzle (Kotovsky & Simon, ; P. J. Reber & Kotovsky, ), we studied the effect on rule discovery of having to plan actions or not in order to reach a goal state. This was done by asking participants to predict legal moves as in implicit learning tasks (Experiment 1) and by asking participants to make legal moves as in problem-solving tasks (Experiment 2). Our hypothesis was that having a specific goal state to reach has a dual effect on rule discovery. The first effect is positive and related to feedback from moves done in order to attain the goal: generalising the results of action and associating them to the conditions in which they were obtained allows discovery of the rule and learning it. The second effect is negative. In attempting to reach a specific goal, participants first tend to reduce the distance that separates the current state from the goal state (hill climbing) and so neglect the kind of exploration that facilitates rule and procedure discovery because this would seem to be a detour from the goal. Results show that having to plan actions improved performance in implicit learning tasks (Experiment 1), yet it impaired performance in problem-solving tasks (Experiment 2). Although implicit learning and problem solving are based on rule discovery, and entail noticing regularities in the material, in both cases, rule discovery processes appear to be task-dependent.     

Caring and the Graduate Student: A Phenomenological Study

Given the increasing numbers of adult learners on university campuses, both as undergraduates and graduates, an understanding of adult learning processes is relevant—indeed essential—to the university's response to this student group. While educators have attempted to understand learners in terms of developmental stages, learning styles, and contextual factors, less attention has focused on the dynamics of helping relationships that attend learning or the importance of caring in the learning process. The purpose of this phenomenological study was to explore and explicate the experience of caring as it relates to graduate education from the perspective of the graduate student. The connection between caring relationships and learning was confirmed. Seven components of the caring were identified. Implications for the university center on attention to relationship skills and the creation of a campus community in which caring relationships are encouraged and nurtured.