Short-term and long-term memory analysis of learning using 2D and 3D educational contents

ABSTRACT

The effect of 2D and 3D educational content learning on memory has been studied using electroencephalography (EEG) brain signal. A hypothesis is set that the 3D materials are better than the 2D materials for learning and memory recall. To test the hypothesis, we proposed a classification system that will predict true or false recall for short-term memory (STM) and long-term memory (LTM) after learning by either 2D or 3D educational contents. For this purpose, EEG brain signals are recorded during learning and testing; the signals are then analysed in the time domain using different types of features in various frequency bands. The features are then fed into a support vector machine (SVM)-based classifier. The experimental results indicate that the learning and memory recall using 2D and 3D contents do not have significant differences for both the STM and the LTM.     

BAM Learning of Nonlinearly Separable Tasks by Using an Asymmetrical Output Function and Reinforcement Learning

Most bidirectional associative memory (BAM) networks use a symmetrical output function for dual fixed-point behavior. In this paper, we show that by introducing an asymmetry parameter into a recently introduced chaotic BAM output function, prior knowledge can be used to momentarily disable desired attractors from memory, hence biasing the search space to improve recall performance. This property allows control of chaotic wandering, favoring given subspaces over others. In addition, reinforcement learning can then enable a dual BAM architecture to store and recall nonlinearly separable patterns. Our results allow the same BAM framework to model three different types of learning: supervised, reinforcement, and unsupervised. This ability is very promising from the cognitive modeling viewpoint. The new BAM model is also useful from an engineering perspective; our simulations results reveal a notable overall increase in BAM learning and recall performances when using a hybrid model with the general regression neural network (GRNN).      

An autoassociative neural network model of paired-associate learning

Hebbian heteroassociative learning is inherently asymmetric. Storing a forward association, from item A to item B, enables recall of B (given A), but does not permit recall of A (given B). Recurrent networks can solve this problem by associating A to B and B back to A. In these recurrent networks, the forward and backward associations can be differentially weighted to account for asymmetries in recall performance. In the special case of equal strength forward and backward weights, these recurrent networks can be modeled as a single autoassociative network where A and B are two parts of a single, stored pattern. We analyze a general, recurrent neural network model of associative memory and examine its ability to fit a rich set of experimental data on human associative learning. The model fits the data significantly better when the forward and backward storage strengths are highly correlated than when they are less correlated. This network-based analysis of associative learning supports the view that associations between symbolic elements are better conceptualized as a blending of two ideas into a single unit than as separately modifiable forward and backward associations linking representations in memory.     

Improved MCMAC with momentum, neighborhood, and averagedtrapezoidal output

An improved modified cerebellar articulation controller (MCMAC) neural control algorithm with better learning and recall processes using momentum, neighborhood learning, and averaged trapezoidal output, is proposed in this paper. The learning and recall processes of MCMAC are investigated using the characteristic surface of MCMAC and the control action exerted in controlling a continuously variable transmission (CVT). Extensive experimental results demonstrate a significant improvement with reduced training time and an extended range of trained MCMAC cells. The improvement in recall process using the averaged trapezoidal output (MCMAC-ATO) are contrasted against the original MCMAC using the square of the Pearson product moment correlation coefficient. Experimental results show that the new recall process has significantly reduced the fluctuations in the control action of the MCMAC and addressed partially the problem associated with the resolution of the MCMAC memory array.     

Hypernetworks: A Molecular Evolutionary Architecture for Cognitive Learning and Memory

Recent interest in human-level intelligence suggests a rethink of the role of machine learning in computational intelligence. We argue that "without cognitive learning the goal of achieving human-level synthetic intelligence is far from completion. Here we review the principles underlying human learning and memory, and identify three of them, i.e., continuity, glocality, and compositionality, as the most fundamental to human-level machine learning. We then propose the recently-developed hypernetwork model as a candidate architecture for cognitive learning and memory. Hypernetworks are a random hypergraph structure higher-order probabilistic relations of data by an evolutionary self-organizing process based on molecular self- assembly. The chemically-based massive interaction for information organization and processing in the molecular hypernetworks, referred to as hyperinteractionism, is contrasted "with the symbolist, connectionist, and dynamicist approaches to mind and intelligence. We demonstrate the generative learning capability of the hypernetworks to simulate linguistic recall memory, visual imagery, and language-vision crossmodal translation based on a video corpus of movies and dramas in a multimodal memory game environment. We also offer prospects for the hyperinteractionistic molecular mind approach to a unified theory of cognitive learning.     

A DNA-based memory with <Emphasis Type="Italic">in vitro</Emphasis> learning and associative recall

A DNA-based memory was implemented with in vitro learning and associative recall.The learning protocol stored the sequences to which it was exposed, and memories were recalled by sequence content through DNA-to-DNA template annealing reactions. Experiments demonstrated that biological DNA could be learned, that sequences similar to the training DNA were recalled correctly, and that unlike sequences were differentiated. Theoretically, the memory has a pattern separation capability that is very large, and can learn long DNA sequences. The learning and recall protocols are massively parallel, as well as simple, inexpensive, and quick. The memory has several potential applications in detection and classification of biological sequences, as well as a massive storage capacity for non-biological data.     

Can differences in learning strategies explain the benefits of learning from static and dynamic visualizations?

The effects of dynamic and static visualizations in understanding physical principles of fish locomotion were investigated. Seventy-five students were assigned to one of three conditions: a text-only, a text with dynamic visualizations, or a text with static visualizations condition. During learning, subjects were asked to think aloud. Learning outcomes were measured by tests assessing verbal factual knowledge, pictorial recall as well as transfer. Learners in the two visualization conditions outperformed those in the text-only condition for transfer and pictorial recall tasks, but not for verbal factual knowledge tasks. Analyses of the think-aloud protocols revealed that learners had generated more inferences in the visualization conditions as opposed to the text-only condition. These results were mirrored by students’ self-reported processing demands. No differences were observable between the dynamic and the static condition concerning any of the learning outcome measures. However, think-aloud protocols revealed an illusion of understanding when learning with dynamic as opposed to static visualizations. Furthermore, learners with static visualizations tended to play the visualizations more often. The results stress the importance of not only using outcome-oriented, but also process-oriented approaches to gain deeper insight into learning strategies when dealing with various instructional materials.     

A bidirectional heteroassociative memory for binary and grey-level patterns

Typical bidirectional associative memories (BAM) use an offline, one-shot learning rule, have poor memory storage capacity, are sensitive to noise, and are subject to spurious steady states during recall. Recent work on BAM has improved network performance in relation to noisy recall and the number of spurious attractors, but at the cost of an increase in BAM complexity. In all cases, the networks can only recall bipolar stimuli and, thus, are of limited use for grey-level pattern recall. In this paper, we introduce a new bidirectional heteroassociative memory model that uses a simple self-convergent iterative learning rule and a new nonlinear output function. As a result, the model can learn online without being subject to overlearning. Our simulation results show that this new model causes fewer spurious attractors when compared to others popular BAM networks, for a comparable performance in terms of tolerance to noise and storage capacity. In addition, the novel output function enables it to learn and recall grey-level patterns in a bidirectional way.     

Using a deep learning network to recognise low back pain in static standing

Abstract

Low back pain (LBP) remains one of the most prevalent musculoskeletal disorders, while algorithms that able to recognise LBP patients from healthy population using balance performance data are rarely seen. In this study, human balance and body sway performance during standing trials were utilised to recognise chronic LBP populations using deep neural networks. To be specific, 44 chronic LBP and healthy individuals performed static standing tasks, while their spine kinematics and centre of pressure were recorded. A deep learning network with long short-term memory units was used for training, prediction and implementation. The performance of the model was evaluated by: (a) overall accuracy, (b) precision, (c) recall, (d) F1 measure, (e) receiver-operating characteristic and (f) area under the curve. Results indicated that deep neural networks could recognise LBP populations with precision up to 97.2% and recall up to 97.2%. Meanwhile, the results showed that the model with the C7 sensor output performed the best.

Practitioner summary: Low back pain (LBP) remains the most common musculoskeletal disorder. In this study, we investigated the feasibility of applying artificial intelligent deep neural network in detecting LBP population from healthy controls with their kinematics data. Results showed a deep learning network can solve the above classification problem with both promising precision and recall performance.     

长文本武侠小说外号识别研究

长文本武侠小说中主人公以侠客和义士为主,人物个性鲜明,外号可以概括人物最显著的特征。传统命名实体识别主要集中在人名、地名、机构名等领域,对于识别外号尚未有相关研究,但作为武侠小说中不可或缺的元素,外号识别对于同义词识别等研究方向具有借鉴意义。鉴于此,该文提出对武侠小说中武侠人名对应的外号的未登录词扩展识别筛选并辅以固定句式法则的识别方法。未登录词扩展识别筛选方法融合了对于左邻字符串的拓展和筛选同时定义了竞争外号子串和候选外号子串等概念,固定句式法则方法是通过外号指示词对观察窗口的候选外号子串进行筛选。经过统计和分类提出了武侠小说高频词表和低频指示字典,用于对竞争外号子串进行筛选。实验证明该文方法可行有效。     

Different identity revelation modes in an online peer-assessment learning environment: Effects on perceptions toward assessors,classroom climate and learning activities

The effects of four different identity revelation modes (three fixed modes: real-name, anonymity, nickname and one dynamic user self-choice mode) on participants’ perceptions toward their assessors, classroom climate, and past experience with the learning activity in which they were engaged were examined. A pretest–posttest quasi-experimental research design was adopted. Eight fifth-grade classes (age 10–11, N = 243) were randomly assigned to four different identity revelation modes in order for them to participate in the study. An online learning system that allows students to contribute to and benefit from the process of question-generation and peer-assessment was adopted. Data analysis confirmed that different identity modes lead participants to view their assessors differently. Specifically, participants assigned to the self-choice and real-name identity revelation modes tended to view their assessors more favorably than those in the anonymity and nickname groups. The empirical significance of the study as well as suggestions for learning system development, instructional implementation and future study are provided.     

Dementia classification using MR imaging and clinical data with voting based machine learning models

Dementia is one of the leading causes of severe cognitive decline, it induces memory loss and impairs the daily life of millions of people worldwide. In this work, we consider the classification of dementia using magnetic resonance (MR) imaging and clinical data with machine learning models. We adapt univariate feature selection in the MR data pre-processing step as a filter-based feature selection. Bagged decision trees are also implemented to estimate the important features for achieving good classification accuracy. Several ensemble learning-based machine learning approaches, namely gradient boosting (GB), extreme gradient boost (XGB), voting-based, and random forest (RF) classifiers, are considered for the diagnosis of dementia. Moreover, we propose voting-based classifiers that train on an ensemble of numerous basic machine learning models, such as the extra trees classifier, RF, GB, and XGB. The implementation of a voting-based approach is one of the important contributions, and the performance of different classifiers are evaluated in terms of precision, accuracy, recall, and F1 score. Moreover, the receiver operating characteristic curve (ROC) and area under the ROC curve (AUC) are used as metrics for comparing these classifiers. Experimental results show that the voting-based classifiers often perform better compared to the RF, GB, and XGB in terms of precision, recall, and accuracy, thereby indicating the promise of differentiating dementia from imaging and clinical data.

     

Case-Based Learning: Predictive Features in Indexing

Interest in psychological experimentation from the Artificial Intelligence community often takes the form of rigorous post-hoc evaluation of completed computer models. Through an example of our own collaborative research, we advocate a different view of how psychology and AI may be mutually relevant, and propose an integrated approach to the study of learning in humans and machines. We begin with the problem of learning appropriate indices for storing and retrieving information from memory. From a planning task perspective, the most useful indices may be those that predict potential problems and access relevant plans in memory, improving the planner's ability to predict and avoid planning failures. This predictive features hypothesis is then supported as a psychological claim, with results showing that such features offer an advantage in terms of the selectivity of reminding because they more distinctively characterize planning situations where differing plans are appropriate.We present a specific case-based model of plan execution, RUNNER, along with its indices for recognizing when to select particular plans—appropriateness conditions—and how these predictive indices serve to enhance learning. We then discuss how this predictive features claim as implemented in the RUNNER model is then tested in a second set of psychological studies. The results show that learning appropriateness conditions results in greater success in recognizing when a past plan is in fact relevant in current processing, and produces more reliable recall of the related information. This form of collaboration has resulted in a unique integration of computational and empirical efforts to create a model of case-based learning.     

汉语基本短语的自动识别

本文应用基于实例的MBL(Memory-Based Learning)学习方法,对汉语中较常见的9种基本短语的边界及类别进行识别,并利用短语内部构成结构和词汇信息对预测中出现的边界歧义和短语类型歧义进行了排歧处理。实验中还比较了在特征向量中加入词汇信息与否对实验结果的影响。实验取得了比较令人满意的结果:对这9种基本短语的识别正确率达到95.2%;召回率达到93.7%。     

The effect of multimedia design types on learners’ recall performances with varying short term memory spans

In this study, the effect of multimedia learning environment designed with two different attention types (focused — split) was investigated on recall performances of learners with different short term memory spans (high — medium — low). The participants were 60 undergraduate students who were presented with either focused attention or split attention multimedia learning materials. First, participants’ short term memory spans were determined by Visual — Aural Digit Span Test-Revised (VADS-B) test. Second, they were separated to three groups as high, medium and low. In 3?×?2 nested ANOVA design, one of the groups studied the multimedia designed in split attention type whereas the other had focused attention type design. As they finished the study task, they were given a recall task, which produced their recall performances. Data were analyzed by Nested ANOVA, t-Test and ANCOVA tests. The findings indicated that multimedia instructional designs were effective on recall performances. Learners showed higher recall performances in the multimedia learning environment in focused attention design. However, no significant difference was observed in learners’ recall performances when their STM spans were taken into account. Significant differences were observed between time spent in studying multimedia.     

Associative processing and processors

Associative memory concerns the concept that one idea may trigger the recall of a different but related idea. Traditional computers, however, rely upon a memory design that stores and retrieves data by its address rather than its content. In such a search, every accessed data word must travel individually between the processing unit and the memory. The simplicity of this retrieval-by-address approach has ensured its success, but has also produced some inherent disadvantages. One is the von Neumann bottleneck, where the memory-access path becomes the limiting factor for system performance. A related disadvantage is the inability to proportionally increase the size of a unit transfer between the memory and the processor as the size of the memory scales up. Associative memory, in contrast, provides a naturally parallel and scalable form of data retrieval for both structured data (e.g. sets, arrays, tables, trees and graphs) and unstructured data (raw text and digitized signals). An associative memory can be easily extended to process the retrieved data in place, thus becoming an associative processor. This extension is merely the capability for writing a value in parallel into selected cells     

Layout attributes and recall

The spatial arrangement of elements such as icons in a computer interface may influence learning the interface. However, the effects of layout organization on users' information processing is relatively little studied so far. The three experiments of this paper examined two attributes of layouts: spatial grouping by proximity and semantic coherence. Learning was assessed by tasks in which 30 participants recalled icon-like items' labels, locations, or both as a series of study-recall trials. The results show that layout organization interacts with task demands. Semantic organization improves recall of labels, and spatial grouping supports recall of locations. When both labels and locations are learned concurrently, the best recall performance is associated with a simultaneously grouped and semantically coherent layout. However, semantic and spatial organization may interact unexpectedly on learning. The findings are discussed from the viewpoint of information chunking in memory processes and interface design.     

Adaptation of the relaxation method for learning in bidirectionalassociative memory

An iterative learning algorithm called PRLAB is described for the discrete bidirectional associative memory (BAM). Guaranteed recall of all training pairs is ensured by PRLAB. The proposed algorithm is significant in many ways. Unlike many existing iterative learning algorithms, PRLAB is not based on the gradient descent technique. It is a novel adaptation from the well-known relaxation method for solving a system of linear inequalities. The algorithm is very fast. Learning 200 random patterns in a 200-200 BAM takes only 20 epochs on the average. PRLAB is highly insensitive to learning parameters and the initial configuration of a BAM. It also offers high scalability for large applications by providing the same high performance when the number of training patterns are increased in proportion to the size of the BAM. An extensive performance analysis of the new learning algorithm is included.     

A comparison of words and icons as external memory aids in an information retrieval task

Enrichment is a process whereby computer-based information is tagged with additional attributes which can be used in an information retrieval system to increase the speed and accuracy of access. In this way, the additional attributes act as external memory aids. Lansdale (1988a) evaluated such a system by looking at the memorability of coloured shapes, placed in different locations on a document, which were used as enrichers in a simple information retrieval task. This paper extends that study to look at memory for labels used in an identical way. Verbal and visual enriching attributes were studied under two conditions: one in which they were assigned to documents automatically by the system, and one in which the users made their own choice. Results indicate a strong trend in which recall was higher when subjects made their own selection of enriching attributes as opposed to having them selected for them. In the comparison of words and icons, there was no evidence that the modalities of the enrichers were a significant factor in recall. Recall performance seems to be primarily related to the 'semantic fit' of the documents and the attributes selected to enrich them. The extent to which this implies potential differences in the utility of visual and verbal methods in future applications is discussed.