Latent Learning Research Articles

LoMAE: Simple Streamlined Low-Level Masked Autoencoders for Robust, Generalized, and Interpretable Low-Dose CT Denoising.

Low-dose computed tomography (LDCT) offers reduced X-ray radiation exposure but at the cost of compromised image quality, characterized by increased noise and artifacts. Recently, transformer models emerged as a promising avenue to enhance LDCT image quality. However, the success of such models relies on a large amount of paired noisy and clean images, which are often scarce in clinical settings. In computer vision and natural language processing, masked autoencoders (MAE) have been recognized as a powerful self-pretraining method for transformers, due to their exceptional capability to extract representative features. However, the original pretraining and fine-tuning design fails to work in low-level vision tasks like denoising. In response to this challenge, we redesign the classical encoder-decoder learning model and facilitate a simple yet effective streamlined low-level vision MAE, referred to as LoMAE, tailored to address the LDCT denoising problem. Moreover, we introduce an MAE-GradCAM method to shed light on the latent learning mechanisms of the MAE/LoMAE. Additionally, we explore the LoMAE's robustness and generability across a variety of noise levels. Experimental findings show that the proposed LoMAE enhances the denoising capabilities of the transformer and substantially reduce their dependency on high-quality, ground-truth data. It also demonstrates remarkable robustness and generalizability over a spectrum of noise levels. In summary, the proposed LoMAE provides promising solutions to the major issues in LDCT including interpretability, ground truth data dependency, and model robustness/generalizability.

3D Multimodal Fusion Network with Disease-induced Joint Learning for Early Alzheimer's Disease Diagnosis.

Multimodal neuroimaging provides complementary information critical for accurate early diagnosis of Alzheimer's disease (AD). However, the inherent variability between multimodal neuroimages hinders the effective fusion of multimodal features. Moreover, achieving reliable and interpretable diagnoses in the field of multimodal fusion remains challenging. To address them, we propose a novel multimodal diagnosis network based on multi-fusion and disease-induced learning (MDL-Net) to enhance early AD diagnosis by efficiently fusing multimodal data. Specifically, MDL-Net proposes a multi-fusion joint learning (MJL) module, which effectively fuses multimodal features and enhances the feature representation from global, local, and latent learning perspectives. MJL consists of three modules, global-aware learning (GAL), local-aware learning (LAL), and outer latent-space learning (LSL) modules. GAL via a self-adaptive Transformer (SAT) learns the global relationships among the modalities. LAL constructs local-aware convolution to learn the local associations. LSL module introduces latent information through outer product operation to further enhance feature representation. MDL-Net integrates the disease-induced region-aware learning (DRL) module via gradient weight to enhance interpretability, which iteratively learns weight matrices to identify AD-related brain regions. We conduct the extensive experiments on public datasets and the results confirm the superiority of our proposed method. Our code will be available at: https://github.com/qzf0320/MDL-Net.

Neural Reinforcement Learning Signals Predict Recovery From Impulse Control Disorder Symptoms in Parkinson’s Disease

BackgroundImpulse control disorders (ICDs) in Parkinson’s disease are associated with a heavy burden on patients and caretakers. While recovery can occur, ICDs persist in many patients despite optimal management. The basis for this interindividual variability in recovery is unclear and poses a major challenge to personalized health care. MethodsWe adopted a computational psychiatry approach and leveraged the longitudinal, prospective Personalized Parkinson Project (136 people with Parkinson’s disease, within 5 years of diagnosis) to combine dopaminergic learning theory–informed functional magnetic resonance imaging with machine learning (at baseline) to predict ICD symptom recovery after 2 years of follow-up. We focused on change in Questionnaire for Impulsive-Compulsive Disorders in Parkinson’s Disease Rating Scale scores in the entire sample regardless of an ICD diagnosis. ResultsGreater reinforcement learning signals during gain trials but not loss trials at baseline, including those in the ventral striatum and medial prefrontal cortex, and the behavioral accuracy score measured while on medication were associated with greater recovery from impulse control symptoms 2 years later. These signals accounted for a unique proportion of the relevant variability over and above that explained by other known factors, such as decreases in dopamine agonist use. ConclusionsOur results provide a proof of principle for combining generative model–based inference of latent learning processes with machine learning–based predictive modeling of variability in clinical symptom recovery trajectories. We showed that reinforcement learning modeling parameters predicted recovery from ICD symptoms in Parkinson’s disease.

„Diplomaţia conferinţelor” sau intelectualizarea politicii externe a RPR sub regimul lui Gheorghiu-Dej

Once Romania came under the influence of the Soviet Union, the new regime established in 1948 was integrated into the diplomatic system created by the Kremlin. The party-state in Bucharest participated in youth festivals, the world women's movement, intellectuals' conferences, the Christian Peace Movement, and particularly the World Peace Congress. All of these events served as propaganda instruments to discourage a more assertive Western anti-Sovietism. Since 1954, Khrushchev initiated a shift in Soviet policy, one focused on peaceful coexistence rather than confrontation between the two formerly irreconcilable political camps. The Romanian People's Republic, along with other state socialist countries, emulated this novel approach in Moscow. The article explores the impact of the new course on Romania’s cultural diplomacy in the late 1950s and the early 1960s. To establish credible diplomatic efforts, there was a need to rethink socialist culture and revalue the pre-communist heritage. This new culture was based on the Leninist principle of “the new culture, national in form, socialist in content”. It signaled that the regime had moved beyond de-Stalinization, and now possessed internal legitimacy that no longer relied on the presence of Soviet troops in the country. However, this socialist culture drew inspiration from postwar autochthonized Bolshevik (Stalinist) nationalism. The medieval tradition rediscovered by Soviet cultural studies (such as Dmitry Likhachev, Viktor Lazarev, or Boris Rybakov) fueled anti-Soviet sentiments in Bucharest, leading, by 1964, to a significant distancing from Moscow. This article traces the diplomatic efforts of the Romanian People's Republic during this period of latent learning of the new rules of the political game within the socialist camp.

Latent learning curve of intellectual capital formation in a public university in central Mexico in the face of the pandemic

The latent growth curve is a technique of structural equations which propose that psychological and social phenomena can be modeled in relationships between factors and indicators. The objective of this work was to establish the differences that reflect the learning of the use of anti-pandemic devices. A longitudinal study was carried out from 2020 to 2024 in a sample of 100 students from a public university in central Mexico who were selected from high school to university. The results show that there are differences, but these are not significant. Such findings agree with studies related to stigma towards anti-pandemic policies. In relation to this state of the art, it is recommended to extend the study to the stigma derived from confinement and distancing in order to establish the learning curve of self-care and prevention. The implications of the study on treatment adherence suggest that stigma would be a latent factor that would be mediating the relationship between the intercept factor and the latent slope factor.

Habituation to onset capture via associative links with contextual information

ABSTRACT Habituation shows organisms’ ability to ignore irrelevant repetitive stimuli, allowing attentional resources to be allocated to the relevant information. Here we investigated the role of contextual information in generating expectations that control habituation to visual onset distractors, expectations that arise from associative learning. Experiment 1 showed that the habituated capture response is restored if the onset distractor is momentarily omitted in the same context of habituation, but not if such omission is experienced in another context. Experiment 2 showed that the pre-exposure to a context in which a distractor is later encountered reduced participants’ capacity to ignore it, thus attesting a latent learning phenomenon affecting habituation of capture. Finally, Experiment 3 showed that ignoring a new distractor is easier in a context in which habituation has already taken place for a different distractor. Taken together, these results attest the context-specific nature of habituation and suggest that learning to ignore salient visual stimuli is based on expectations encoded through associative learning.

Unsupervised Feature Selection with Latent Relationship Penalty Term

With the exponential growth of high dimensional unlabeled data, unsupervised feature selection (UFS) has attracted considerable attention due to its excellent performance in machine learning. Existing UFS methods implicitly assigned the same attribute score to each sample, which disregarded the distinctiveness of features and weakened the clustering performance of UFS methods to some extent. To alleviate these issues, a novel UFS method is proposed, named unsupervised feature selection with latent relationship penalty term (LRPFS). Firstly, latent learning is innovatively designed by assigning explicitly an attribute score to each sample according to its unique importance in clustering results. With this strategy, the inevitable noise interference can be removed effectively while retaining the intrinsic structure of data samples. Secondly, an appropriate sparse model is incorporated into the penalty term to further optimize its roles as follows: (1) It imposes potential constraints on the feature matrix to guarantee the uniqueness of the solution. (2) The interconnection between data instances is established by a pairwise relationship situation. Extensive experiments on benchmark datasets demonstrate that the proposed method is superior to relevant state-of-the-art algorithms with an average improvement of 10.17% in terms of accuracy.

Fruit bats adjust their decision-making process according to environmental dynamics

One of the main functions of behavioral plasticity lies in the ability to contend with dynamic environments. Indeed, while numerous studies have shown that animals adapt their behavior to the environment, how they adapt their latent learning and decision strategies to changes in the environment is less understood. Here, we used a controlled experiment to examine the bats’ ability to adjust their decision strategy according to the environmental dynamics. Twenty-five Egyptian fruit bats were placed individually in either a stable or a volatile environment for four consecutive nights. In the stable environment, two feeders offered food, each with a different reward probability (0.2 vs. 0.8) that remained fixed over two nights and were then switched, while in the volatile environment, the positions of the more and the less rewarding feeders were changed every hour. We then fit two alternative commonly used models namely, reinforcement learning and win-stay-lose-shift strategies to the bats' behavior. We found that while the bats adapted their decision-making strategy to the environmental dynamics, they seemed to be limited in their responses based on natural priors. Namely, when the environment had changed slowly, at a rate that is natural for these bats, they seemed to rely on reinforcement learning and their performance was nearly optimal, but when the experimental environment changed much faster than in the natural environment, the bats stopped learning and switched to a random decision-making strategy. Together, these findings exemplify both the bats’ decision-making plasticity as well as its natural limitations.

Finite-Sample Analysis of Deep CCA-Based Unsupervised Post-Nonlinear Multimodal Learning.

Canonical correlation analysis (CCA) has been essential in unsupervised multimodal/multiview latent representation learning and data fusion. Classic CCA extracts shared information from multiple modalities of data using linear transformations. In recent years, deep neural networks-based nonlinear feature extractors were combined with CCA to come up with new variants, namely the ``DeepCCA'' line of work. These approaches were shown to have enhanced performance in many applications. However, theoretical supports of DeepCCA are often lacking. To address this challenge, the recent work of Lyu and Fu (2020) showed that, under a reasonable postnonlinear generative model, a carefully designed DeepCCA criterion provably removes unknown distortions in data generation and identifies the shared information across modalities. Nonetheless, a critical assumption used by Lyu and Fu (2020) for identifiability analysis was that unlimited data is available, which is unrealistic. This brief paper puts forth a finite-sample analysis of the DeepCCA method by Lyu and Fu (2020). The main result is that the finite-sample version of the method can still estimate the shared information with a guaranteed accuracy when the number of samples is sufficiently large. Our analytical approach is a nontrivial integration of statistical learning, numerical differentiation, and robust system identification, which may be of interest beyond the scope of DeepCCA and benefit other unsupervised learning paradigms.

Dynamic technological capabilities in leading European R&D firms: a panel data analysis with structural equations

ABSTRACT The deployment of R&D-driven dynamic technological capabilities (DTCs) is related to firm performance and represents a critical point in global socio-economic development. This is an interesting topic for EU economic, political and business actors confronted with increasingly changing environments. This study analyzed the influence of DTCs on firm performance using a longitudinal approach based on a structural equation modelling technique with a panel dataset with 268 observations from a sample of 67 leading European firms in the Information and Communication Technology (ICT) industry. The results show that firm performance is positively influenced by the deployment of DTCs, and this effect is significant, at least in the subsequent year's performance. These results shed light on the latent learning effect in the development of R&D-driven DTCs in ICT firms.

Ants combine object affordance with latent learning to make efficient foraging decisions

The affordance of an object refers to its functional properties. For example, a bowl has the affordance of holding water, but a sieve does not. Here, we report that ants learn the affordance of a novel object without this attribute being rewarded, and use the memory of this affordance to avoid predicted, but never experienced, crowding. Ants were trained to feeders, which could support either only one ant or many. Two feeders were encountered, each of identical design but differently scented. After training, on the outward journey, half the ants encounter nestmates, which had fed on food matching one of the training feeders. Encountering returning nestmates reduced preference for the feeder matching the scent of the encountered nestmates, but only for ants trained on a limited-access feeder; ants trained on an unlimited feeder were unaffected. In other words, only if ants know that the food access is limited, and receive information that this feeder is heavily visited, do they reduce their preference for this feeder. To achieve this, the ants had to combine memories of the feeders' affordance with the presence of nestmates. Then they had to use semantic knowledge that restricted food access combined with nestmate presence predicts a likelihood of crowding, or a rule such as "if the food is restricted and there are nestmates on the path, go to another food source." Regardless of the mechanism, these results demonstrate that ants latently learn the affordance of their surroundings, an unexpected cognitive ability for an invertebrate.

Chronic dietary exposure to arsenic at environmentally relevant concentrations impairs cognitive performance in adult zebrafish (Danio rerio) via oxidative stress and dopaminergic dysfunction

The current study was designed to evaluate the effects of chronic dietary arsenic exposure on the cognitive performance of adult zebrafish and uncover probable pathways by which arsenic mediates such neurotoxic effects. Adult zebrafish were treated with 3 different dietary arsenic concentrations (30, 60, and 100 μg/g dry weight (dw), as arsenite) in addition to control for 60 days. A latent learning paradigm, which employs a complex maze, was used to assess the cognitive performance of fish. Our results demonstrated that dietary treatment with arsenic, especially at medium (60 μg/g dw) and high (100 μg/g dw) exposure dose levels, significantly impaired the performance of fish in various latent learning tasks evaluated in the present study. Concomitant with cognitive dysfunction, chronic dietary exposure to arsenic was also found to increase arsenic accumulation and dopamine levels, and induce oxidative stress (reduced thiol redox, increased lipid peroxidation and expression of antioxidant enzyme genes) in the brain of zebrafish in a dose-dependent manner. Dopaminergic system in the brain is known to play a critical role in regulating cognitive behaviours in fish, and our observations suggested that chronic dietary treatment with medium and high arsenic doses leads to significant alterations in the expression of genes involved in dopamine signaling (dopamine receptors), synthesis (thyroxine hydroxylase) and metabolism (monoamine oxidase) in the zebrafish brain. Moreover, we also recorded significant downregulation of genes such as the brain-derived neurotrophic factor (BDNF) and ectonucleotidases (entpd2_mg, entpd2_mq, and 5′-nucleotidase), which are critical for learning and memory functions, in the zebrafish brain following chronic dietary exposure to arsenic. Overall, the present study suggests that chronic environmentally relevant dietary exposure to arsenic can impair the cognitive performance in zebrafish, essentially by inducing oxidative stress and disrupting the dopaminergic neurotransmission in the brain.

Learning data teaching strategies via knowledge tracing

Teaching plays a fundamental role in human learning. Typically, a human teaching strategy involves assessing a student’s knowledge progress for tailoring the teaching materials to enhance the learning progress. A human teacher can achieve this by tracing a student’s knowledge over essential learning concepts in a task. Albeit, such a teaching strategy is not well exploited yet in machine learning as current machine teaching methods tend to directly assess the progress of individual training samples without paying attention to the underlying learning concepts in a learning task. In this paper, we propose a novel method, called Knowledge Augmented Data Teaching (KADT), which can optimize a data teaching strategy for a student model by tracing its knowledge progress over multiple learning concepts in a learning task. Specifically, the KADT method incorporates a knowledge tracing model to dynamically capture the knowledge progress of a student model in terms of latent learning concepts. We further develop an attention-pooling mechanism to distill knowledge representations of a student model with respect to class labels, which enables to develop a data teaching strategy on critical training samples. We have evaluated the performance of the KADT method on four different machine learning tasks, including knowledge tracing, sentiment analysis, movie recommendation, and image classification. The KADT method consistently outperforms the state-of-the-art methods on all these tasks.

Design and development process of a digital game to acquire preschoolers’ self-care skills: a model practice

This study aims to clearly describe how to design and develop a digital game with a concrete example for preschoolers to acquire Self-Care Skills (SCS). The study is in the model of design-based developmental research and includes such participants as subject-matter experts, practitioner teachers, and preschoolers in the development steps. RETRO game development model was adapted and followed in the game development process by integrating Digital Game-Based Learning (DGBL) approach with entertainment game development. Each step of the model involves gathering, analyzing data and yielding informative findings that guide the next step. The views of participants indicate that some behaviors in SCS can be cognitively acquired and internalized by the game with latent learning. While realizing learning outcomes, the developed game succeeded in motivation, engagement and design features as in commercial entertainment games. As the model recommends, if all the bases are covered, the design and development process may be executed right. Since the study reports the entire process of the concrete example, it can help guide novice digital game developers (can be a teacher, a trainer as well) repeat the game development or develop a new one based on DGBL. Keywords: Digital games, digital game-based learning, early year’s education, mobile games, self-care skills.

Latent Learning Disabilities in the Classroom: Interpreting Children's Learning Styles Identified by Memory Recalls

If a learning disability is not defined, it can be said that primary school children show the features of adaptable students within the learning atmosphere of the classroom. Most of the time, teachers think that they teach and their students learn easily. However, studies on children’s memory show that the learning process gets abstract when the number of stimulants in the learning environment increase. Children especially have difficulties when they use more than one sensory memory, like seeing, hearing and touching, at the same time and it gets harder to remember things they learned. In this study, it is aimed to investigate the learning profiles related to the visual, auditory, kinesthetic & tactile and combined (visual+auditory+kinesthetic) memories of primary school children and to define whether learning results related to the combined memory cause a learning disability in terms of remembering. According to the results of the study, it has come out that a considerable amount of children have difficulties while learning with the combined memory. However, their teachers and probably also their parents are not aware of this situation. This result makes teachers consider that they have to help their students out individually in the learning environment in order to make students’ learning easier.

Effectiveness of learning and growth performance metrics in the Nepalese telecommunications industry for organizational success

The primary use of financial-based performance metrics to assess an organization’s success might be misleading. The application of non-financial performance metrics could improve organizational success and longevity. This study aimed to examine the effectiveness of learning and growth performance metrics for organizational success in the Nepalese telecommunication industry. The quantitative research approach was utilized for collecting, presenting, and analyzing data obtained during a survey. The two major telecommunications service providers in Nepal, Ncell and Nepal Telecom, were taken as sample organizations, and their employees were the study’s respondents. The study revealed that two latent learning and growth performance metrics, namely ‘organizational culture and alignment’ having seven observable variables (β = 0.229, t = 3.419, p < .05) and ‘information capital’ having four observable variables (β = 0.079, t = 1.193, p < .05) were significant for organizational success. In contrast, one latent metric, ‘human resources’ having seven observable variables (β = 0.047, t = 0.708, p > .05), was insignificant. The overall explanation of the observed non-financial performance metrics to the organizational success of the Nepalese telecommunication industry was approximately 6%. A better learning and growth environment helps an organization generate, acquire, share, and integrate information to build resources and capabilities. In addition, non-financial performance metrics help organizations connect business performance with strategy, allowing them to be competitive.

«Уроки выходного дня» как методический прием формирования умений устной речи на русском языке иностранных студентов в рамках ситуативно-обусловленного латентного обучения

The aim of the study is to show the expediency of using a situated approach to teaching Russian as a foreign language (RFL) to foreign students in conditions of latent learning. The paper is the result of a thorough understanding of the practical experience of RFL teaching at the preparatory faculty of the Udmurt State University. The paper analyses the types of speech situations and educational speech situations in terms of the advantages and disadvantages of using them in the educational process; considers the issue of a successful use of situated learning in RFL teaching. Main attention is focused on the possibility of implementing situated learning in real communication in a latent, covert, implicit manner. Scientific novelty of the study lies in developing the methodological technique “Weekend Lessons” in order to form oral speech skills among preparatory faculty students. As a result, the main levels of speaking skills formation using the methodological technique “Weekend Lessons”, the stages of implementation of the specified methodological technique have been characterised, recommendations for its further application in the practice of RFL teaching have been provided.

Dopamine Release Impairments Accompany Locomotor and Cognitive Deficiencies in Rotenone-Treated Parkinson's Disease Model Zebrafish.

In this work, we carried out neurochemical and behavioral analysis of zebrafish (Danio rerio) treated with rotenone, an agent used to chemically induce a syndrome resembling Parkinson's disease (PD). Dopamine release, measured with fast-scan cyclic voltammetry (FSCV) at carbon-fiber electrodes in acutely harvested whole brains, was about 30% of that found in controls. Uptake, represented by the first order rate constant (k) and the half-life (t1/2) determined by nonlinear regression modeling of the stimulated release plots, was also diminished. Behavioral analysis revealed that rotenone treatment increased the time required for zebrafish to reach a reward within a maze by more than 50% and caused fish to select the wrong pathway, suggesting that latent learning was impaired. Additionally, zebrafish treated with rotenone suffered from diminished locomotor activity, swimming shorter distances with lower mean velocity and acceleration. Thus, the neurochemical and behavioral approaches, as applied, were able to resolve rotenone-induced differences in key parameters. This approach may be effective for screening therapies in this and other models of neurodegeneration.

Impaired Dopamine Release and Latent Learning in Alzheimer's Disease Model Zebrafish.

Alzheimer's disease (AD) is a progressive, fatal, neurodegenerative disorder for which only treatments of limited efficacy are available. Despite early mentions of dementia in the ancient literature and the first patient diagnosed in 1906, the underlying causes of AD are not well understood. This study examined the possible role of dopamine, a neurotransmitter that is involved in cognitive and motor function, in AD. We treated adult zebrafish (Danio rerio) with okadaic acid (OKA) to model AD and assessed the resulting behavioral and neurochemical changes. We then employed a latent learning paradigm to assess cognitive and motor function followed by neurochemical analysis with fast-scan cyclic voltammetry (FSCV) at carbon fiber microelectrodes to measure the electrically stimulated dopamine release. The behavioral assay showed that OKA treatment caused fish to have lower motivation to reach the goal chamber, resulting in impeded learning and decreased locomotor activity compared to controls. Our voltammetric measurements revealed that the peak dopamine overflow in OKA-treated fish was about one-third of that measured in controls. These findings highlight the profound neurochemical changes that may occur in AD. Furthermore, they demonstrate that applying the latent learning paradigm and FSCV to zebrafish is a promising tool for future neurochemical studies and may be useful for screening drugs for the treatment of AD.

TEKNIK KOMUNIKASI DAN TEKNIK PEMBELAJARAN ANAK BERKEBUTUHAN KHUSUS PADA MASA PANDEMI COVID 19

This study aims to describe communication techniques and learning techniques for Children with Special Needs during the Covid19 pandemic at KTK BSS Giwangan Yogyakarta. The research used descriptive qualitative research. The qualitative approach was chosen because this study used social interaction to obtain data from natural data sources. The data source of this research is Children with Special Needs (ABK) or Special Needs at KTK BSS Giwangan Yogyakarta in kindergarten and elementary school levels. From this research, it was found that in KTK BSS Giwangan there are three types of children with special needs, namely speech impaired, autism and learning difficulties. Communication techniques used are: 1) Using private learning methods (one teacher one child); 2) Make eye contact with the child you are talking to; 3) Cultivating a sense of empathy in conversation; 4) Discuss specific and clear topics and 5) Be patient in waiting for answers from children. Different learning techniques from other children. The learning technique used is the learning technique used for the speech impaired using the Indonesian Sign Language System (SIBI) and the Indonesian Sign Language (BISINDO). For children with Autism constraints using 1) Modeling; 2) Latent Learning 3) Give positive praise and 4) Divide all activities step by step. Learning Difficulties using 1) Multisensory Method; 2) Linguistic Method and 3) Glass Analysis