Sequential Learning Process Research Articles

Unveiling user interests: A deep user interest exploration network for sequential location recommendation

Sequential location recommendation, also known as next location prediction, is a crucial task aiming to compute the likelihood of a user visiting a certain point of interest (POI). It has significant applications in route planning and location-based advertisements. Despite the utilization of geographical information, existing methods often adopt a rigidly sequential prediction mode, focusing solely on predicting the next destination (i.e., where to go), which limits their ability to effectively model user trajectories in terms of predicting from where the user came. Moreover, these methods struggle to capture the evolving and deep user interests along their trajectory. To deal with these problems, we propose a novel approach called Deep User Interest Exploration Network (DUIEN) for sequential location recommendation. DUIEN incorporates two essential components to enhance recommendation accuracy. First, we employ a bidirectional neural network with the Cloze objective, enabling it to predict random masked POIs within user trajectories. This approach aids the model in gaining a better understanding of the context and dynamics of user trajectories. Second, we introduce a user interest exploration layer that seamlessly integrates into the sequential learning process. This layer effectively utilizes tag information associated with POIs and learns the intricate interplay between user interests and their trajectories using a deep user interest exploration method. In our experiments, DUIEN outperforms state-of-the-art sequential location recommendation models, demonstrating the effectiveness of the designed update mechanisms for deep user interest exploration. This showcases the potential of our approach to significantly enhance the accuracy of sequential location recommendation.

Impact of Nutrition Education on Physical Fitness in School-Aged Children in Japan

Purpose: The aim of the study was to investigate the impact of nutrition education on physical fitness in school-aged children in japan.
 Methodology: This study adopted a desk methodology. A desk study research design is commonly known as secondary data collection. This is basically collecting data from existing resources preferably because of its low cost advantage as compared to a field research. Our current study looked into already published studies and reports as the data was easily accessed through online journals and libraries.
 Findings: The impact of nutrition education on physical fitness in school-aged children in Japan has been a focus of research and policy efforts. Several studies conducted in Japan have demonstrated the positive effects of nutrition education on enhancing the physical fitness of children. These programs have emphasized the importance of balanced diets, nutrition knowledge, and healthy eating habits. Additionally, they have highlighted the role of schools and families in promoting physical activity and nutritious choices among children. Overall, the evidence from Japan underscores the significance of comprehensive nutrition education in improving the physical fitness and overall health of school-aged children in the country.
 Unique Contribution to Theory, Practice and Policy: Health Belief Model (HBM), Social Cognitive Theory (SCT) and Theory of Planned Behavior (TPB) may be used to anchor future studies on impact of nutrition education on physical fitness in school-aged children in japan. Educational institutions should integrate nutrition education into their curriculum from an early age and ensure that it is part of a continuous and sequential learning process. Policymakers should consider making nutrition education mandatory in school curricula, with standardized guidelines for content and delivery.

Accurate Prediction of Three-Body Intermolecular Interactions via Electron Deformation Density-Based Machine Learning.

This work extends the electron deformation density-based descriptor, originally developed in the electron deformation density-based interaction energy machine learning (EDDIE-ML) algorithm to predict dimer interaction energies, to the prediction of three-body interactions in trimers. Using a sequential learning process to select the training data, the resulting Gaussian process regression (GPR) model predicts the three-body interaction energy within 0.2 kcal mol-1 of the SRS-MP2/cc-pVTZ reference values for the 3B69 and S22-3 trimer data sets. A hybrid kernel function is introduced, which combines contributions from the average and individual atomic environments, allowing the total trimer interaction energy to be predicted in addition to the three-body contribution using the same descriptor. To extend the range and diversity of trimer interaction energies available in the literature, a new data set based on a protein-ligand crystal structure is introduced, consisting of 509 structures of a central ligand with two protein fragments. Benchmark calculations are provided for the new data set, which contains significantly larger molecular interactions than current databases in the literature in addition to charged fragments. Compared to density funtional theory (DFT)- and wavefunction-based methods for calculating the three-body interaction energy, our model makes predictions in a significantly shorter time frame by reducing the number of required SCF calculations from 7 to 4 performed at the PBE0 level of theory, showcasing the utility and efficiency of our Δ-ML method particularly when applied to larger systems.

Project EARTH: Lessons from 10 Years of Teaching Public Health Skills for Resource-Limited Settings.

The College of Public Health at East Tennessee State University started a program in 2011 to teach the skills needed to protect and promote health and well-being in resource-limited settings. The need to provide public health services in resource-limited settings exists in both wilderness and isolated settings and when a disaster disrupts basic societal infrastructure. In these settings, lives may depend on the ability to provide water, sanitation, hygiene, shelter, first aid, and other basic services. Over the last decade, the college expanded the program considerably into what is now known as Project EARTH (Employing Available Resources to Transform Health) that now includes several different academic courses as well as programs designed to develop innovative solutions to address the needs of people in resource-limited settings. Working in a resource-limited setting requires effectively utilizing locally available resources to improve and protect people's health and well-being. Project EARTH focuses on teaching students to design and create specific products for these situations while progressively honing those cross-cutting skills necessary to work effectively in these settings-notably teamwork, creativity, and resilience. To this end, Project EARTH implements a sequential learning process that includes significant hands-on training and simulatedexperiences with debriefing opportunities at the end of each activity. Project EARTH may serve as a useful model for others considering a similar training program.

Incremental Embedding Learning via Zero-Shot Translation

Modern deep learning methods have achieved great success in machine learning and computer vision fields by learning a set of pre-defined datasets. Howerver, these methods perform unsatisfactorily when applied into real-world situations. The reason of this phenomenon is that learning new tasks leads the trained model quickly forget the knowledge of old tasks, which is referred to as catastrophic forgetting. Current state-of-the-art incremental learning methods tackle catastrophic forgetting problem in traditional classification networks and ignore the problem existing in embedding networks, which are the basic networks for image retrieval, face recognition, zero-shot learning, etc. Different from traditional incremental classification networks, the semantic gap between the embedding spaces of two adjacent tasks is the main challenge for embedding networks under incremental learning setting. Thus, we propose a novel class-incremental method for embedding network, named as zero-shot translation class-incremental method (ZSTCI), which leverages zero-shot translation to estimate and compensate the semantic gap without any exemplars. Then, we try to learn a unified representation for two adjacent tasks in sequential learning process, which captures the relationships of previous classes and current classes precisely. In addition, ZSTCI can easily be combined with existing regularization-based incremental learning methods to further improve performance of embedding networks. We conduct extensive experiments on CUB-200-2011 and CIFAR100, and the experiment results prove the effectiveness of our method. The code of our method has been released in https://github.com/Drkun/ZSTCI.

Multimodal Learning Based Spatial Relation Identification

Spatial Relation identification is one of the integral parts of Spatial Information Retrieval. It deals with identifying the spatially related objects in view of their physical orientation or placement with respect to each other. The concept is widely used in many fields such as Robotics, Image Caption Generation and many more such areas. In this work the focus is to gather information from multiple modalities such as Image and its corresponding Text so as to strengthen the learning process for the identification of Spatial Relation pairs from a given text. Two different multimodal approaches are proposed in this work. In the first approach, information is explored as a sequential learning process where the individual Spatial Roles are identified as connected entities, which makes the Spatial Relation retrieval easy and efficient enough. To counter the small size of the dataset along with necessity to avoid overfitting, an efficient backward propagation based Neural Network was used to classify the candidate roles and the relations. The feature selection was different for all the classification tasks. Building on the selected feature from the first approach, the second approach uses a transfer learning method that utilizes an existing image caption generation model to retrieve the vital topic based information from image which is then used for the task. Thereby both approaches used information from two modalities which are further used to train the system in the respective approach. The model achieves state-of-the-art performance in terms of Precision for two of the Spatial Roles identification. This validates the advantage of using multimodal learning when compared with other partial-multimodal processes.

Incremental learning based multi-domain adaptation for object detection

Cross-domain object detection uses knowledge from source domain tasks to enhance the object detection in target domain. It can reduce the workload of data annotations in the new domain and significantly improve the adaptation ability of the network. We consider a more realistic transfer scenario, that is, our target domain samples cannot be obtained at the same time so that learners cannot forget the old domain knowledge while learning the new domain knowledge, which is consistent with the basic assumption of incremental learning. However, the existing methods do not consider the sequential learning process of the multiple target domains which may cause performance degradation. To address this issue, we propose a novel multi-domain adaptation method for object detection based on incremental learning. Specifically, the incremental learning network saves the knowledge of multiple domains and makes the model to fuse the knowledge of different domains during the training effectively. To make it work better, the progressive training strategy is proposed to make the model gradually adapt to multiple domains. Moreover, we use a multi-level feature alignment module to ensure that domain alignment is realized on features at various levels. We perform experiments on two sets of 6 datasets, which demonstrate that our model can effectively solve the problem of domain knowledge forgetting in multi-target domain adaptation and significantly improve the detection accuracy in each domain.

Internet of emotional people: Towards continual affective computing cross cultures via audiovisual signals

Despite considerable advances achieved in affective computing over the past decade, the related learning paradigms still remain in an isolated fashion, i.e., models are often designed and developed task-dependently. Nevertheless, with the inherently heterogeneous and dynamic property of multiple tasks, we inevitably face several challenges, such as the implementation feasibility, when dealing with the growing number of new tasks of interest. For this reason, in this study, we endeavour to shed some fresh light on shifting the conventional isolated affective computing into a lifelong learning paradigm, namely continual affective computing. As the first tentative work in audio and video domains, we explore the lifelong learning algorithm of elastic weight consolidation for this benchmark work, in an application of well-established audiovisual emotion recognition in a cross-culture scenario, i.e., French and German emotion recognition. To evaluate the feasibility and effectiveness of the introduced lifelong learning, we perform extensive experiments across the RECOLA and SEWA databases. The empirical results show that the implemented lifelong learning approach remarkably outperforms other baselines in most cases, and is even competitive to the joint training process in some cases, indicating its capability when handling the sequential learning process with multiple tasks.

The Extended Congress: Reimagining scientific meetings after the COVID-19 pandemic.

This article was migrated. The article was marked as recommended. The COVID-19 pandemic has profoundly impacted the medical meetings planned for 2020. This health crisis has caused the cancellation, postponement or a pivot in educational design to virtual formats. In the latter case, the format for virtual meetings has remained very similar to the cancelled face-to-face meeting, by using primarily web conferencing systems. This article intends to start a dialogue with the medical education and events community about possible delivery formats. Among them, the concept of an "Extended Congress" is introduced. The extended congress uses the extension of time, space and languages to a scientific meeting. It aims to: 1) unleash the reach of traditional meetings through the use of technology to access larger audiences in different languages, across a country and internationally, with local leaders to help interpret the knowledge and localize it, and 2) to improve knowledge translation into practice through a sequential and active learning process. An ongoing example is described as a proof of concept: the Latin American Peritoneal Dialysis Extended Congress attracted 774 remote participants from over 20 countries, 93% of whom were paid registrants. Initially designed as a hybrid (live plus remote) event scheduled for March 2020, it had to be reframed as a remote only meeting due to the COVID-19 pandemic, thus protecting the health of members while providing continued value to the organization and attendees of the event. With this experience in mind, the authors are currently designing programs in the United States, through collaboration with the University of Virginia Office of Continuing Medical Education. In summary, the design of meetings can better utilize and integrate technology and reach larger audiences with a blend of formats. Those organizations that adapt more quickly to offer these events will concentrate more of the share, as seen with the adoption of technology by other industries.

Deep Learning–Based Survival Analysis Identified Associations Between Molecular Subtype and Optimal Adjuvant Treatment of Patients With Gastric Cancer

PurposeGastric cancer (GC) is the third-leading cause of cancer-related deaths. Several pivotal clinical trials of adjuvant treatments were performed during the previous decade; however, the optimal regimen for adjuvant treatment of GC remains controversial.Patients and MethodsWe developed a novel deep learning–based survival model (survival recurrent network [SRN]) in patients with GC by including all available clinical and pathologic data and treatment regimens. This model uses time-sequential data only in the training step, and upon being trained, it receives the initial data from the first visit and then sequentially predicts the outcome at each time point until it reaches 5 years. In total, 1,190 patients from three cohorts (the Asian Cancer Research Group cohort, n = 300; the fluorouracil, leucovorin, and radiotherapy cohort, n = 432; and the Adjuvant Chemoradiation Therapy in Stomach Cancer cohort, n = 458) were included in the analysis. In addition, we added Asian Cancer Research Group molecular classifications into the prediction model. SRN simulated the sequential learning process of clinicians in the outpatient clinic using a recurrent neural network and time-sequential outcome data.ResultsThe mean area under the receiver operating characteristics curve was 0.92 ± 0.049 at the fifth year. The SRN demonstrated that GC with a mesenchymal subtype should elicit a more risk-adapted postoperative treatment strategy as a result of its high recurrence rate. In addition, the SRN found that GCs with microsatellite instability and GCs of the papillary type exhibited significantly more favorable survival outcomes after capecitabine plus cisplatin chemotherapy alone.ConclusionOur SRN predicted survival at a high rate, reaching 92% at postoperative year 5. Our findings suggest that SRN-based clinical trials or risk-adapted adjuvant trials could be considered for patients with GC to investigate more individualized adjuvant treatments after curative gastrectomy.

Inverse Tone Mapping Operator Using Sequential Deep Neural Networks Based on the Human Visual System

Conventional digital displays show images with much less dynamic range than that of human visual perception. High-dynamic range (HDR) displays are being developed for viewing images with higher dynamic range than that of conventional low-dynamic range (LDR) images. However, to view existing LDR images on HDR displays, an inverse tone mapping operator (ITMO), which is a process to extend the dynamic range of LDR images, is required. In this paper, we propose an adaptive ITMO by effectively learning the differences between LDR and HDR images using a sequential learning process: dynamic range learning followed by color difference learning. Our proposed method enables visualization of colors similar to real-world colors better than conventional ITMOs by learning color differences based on the human visual system properties. For the objective comparison, seven different evaluation metrics optimized for HDR image evaluation were used. Our method resulted in 10%–25% improved HDR-visual difference predictor-2.2 values over those of other ITMOs. Other metrics also demonstrated the superior performance of our method. For the subjective comparison, eight human observers evaluated the estimated HDR images in terms of color appearance and overall preferences. Our method received an average 4.7 out of 5 score, whereas other ITMOs received below 3.5 scores in the evaluation of color appearance. The objective and subjective evaluations’ results showed that our proposed method outperformed the conventional ITMOs in estimating the dynamic range and color appearance of ground-truth HDR images.

Extreme learning machine based transfer learning algorithms: A survey

Extreme learning machine (ELM) has been increasingly popular in the field of transfer learning (TL) due to its simplicity, training speed and ease of use in online sequential learning process. This paper critically examines transfer learning algorithms formulated with ELM technique and provides state of the art knowledge to expedite the learning process ELM based TL algorithms. As this article discusses available ELM based TL algorithm in detail, it provides a holistic overview of current literature, serves as a starting point for new researchers in ELM based TL algorithms and facilitates identification of future research direction in concise manner.

A simple computational algorithm of model-based choice preference.

A broadly used computational framework posits that two learning systems operate in parallel during the learning of choice preferences-namely, the model-free and model-based reinforcement-learning systems. In this study, we examined another possibility, through which model-free learning is the basic system and model-based information is its modulator. Accordingly, we proposed several modified versions of a temporal-difference learning model to explain the choice-learning process. Using the two-stage decision task developed by Daw, Gershman, Seymour, Dayan, and Dolan (2011), we compared their original computational model, which assumes a parallel learning process, and our proposed models, which assume a sequential learning process. Choice data from 23 participants showed a better fit with the proposed models. More specifically, the proposed eligibility adjustment model, which assumes that the environmental model can weight the degree of the eligibility trace, can explain choices better under both model-free and model-based controls and has a simpler computational algorithm than the original model. In addition, the forgetting learning model and its variation, which assume changes in the values of unchosen actions, substantially improved the fits to the data. Overall, we show that a hybrid computational model best fits the data. The parameters used in this model succeed in capturing individual tendencies with respect to both model use in learning and exploration behavior. This computational model provides novel insights into learning with interacting model-free and model-based components.

M-estimator-based online sequential extreme learning machine for predicting chaotic time series with outliers

An M-estimator-based online sequential extreme learning machine (M-OSELM) is proposed to predict chaotic time series with outliers. The M-OSELM develops from the online sequential extreme learning machine (OSELM) algorithm and retains the same excellent sequential learning ability as OSELM, but replaces the conventional least-squares cost function with a robust M-estimator-based cost function to enhance the robustness of the model to outliers. By minimizing the M-estimator-based cost function, the possible outliers are prevented from entering the model’s output weights updating scheme. Meanwhile, in the sequential learning process of M-OSELM, a sequential parameter estimation approach based on error sliding window is introduced to estimate the threshold value of the M-estimator function for online outlier detection. Thanks to the built-in median operation and sliding window strategy, this approach is efficient to provide a stable estimator continuously without high computational costs, and then the potential outliers can be effectively detected. Simulation results show that the proposed M-OSELM has an excellent immunity to outliers and can always achieve better performance than its counterparts for prediction of chaotic time series when the training dataset contains outliers, ensuring at the same time all benefits of an online sequential approach.

A SEQUENTIAL LEARNING METHOD FOR DETECTION AND CLASSIFICATION OF EXUDATES IN RETINAL IMAGES TO ASSESS DIABETIC RETINOPATHY

One of the greatest concerns to the personnel in the current health care sector is the severe progression of diabetes. People can often have diabetes and be completely unaware as the symptoms seem harmless when they are seen on their own. Diabetic retinopathy (DR) is an eye disease that is associated with long-standing diabetes. Retinopathy can occur with all types of diabetes and can lead to blindness if left untreated. The conventional method followed by ophthalmologists is the regular testing of the retina. As this method takes time and energy of the ophthalmologists, a new feature-based automated technique for classification and detection of exudates in color fundus image is proposed in this paper. This method reduces the work of the professionals while examining every fundus image rather than only on abnormal image. The exudates are detected from the color fundus image by applying a few pre-processing techniques that remove the optic disk and similar blood vessels using morphological operations. The pre-processed image was then applied for feature extraction and these features were utilized for classification purpose. In this paper, a novel classification technique such as self-adaptive resource allocation network (SRAN) and meta-cognitive neural network (McNN) classifier is employed for classification of images as exudates, their severity and nonexudates. SRAN classifier makes use of self-adaptive thresholds to choose the appropriate training samples and removes the redundant samples to prevent over-training. These selected samples are availed to improve the classification performance. McNN classifier employs human-like meta-cognition to regulate the sequential learning process. The meta-cognitive component controls the learning process in the cognitive component by deciding what-to-learn, when-to-learn and how-to-learn. It is therefore evident that the implementation of human meta-cognitive learning principle improves efficient learning.

Embedding Responses in Spontaneous Neural Activity Shaped through Sequential Learning

Recent experimental measurements have demonstrated that spontaneous neural activity in the absence of explicit external stimuli has remarkable spatiotemporal structure. This spontaneous activity has also been shown to play a key role in the response to external stimuli. To better understand this role, we proposed a viewpoint, “memories-as-bifurcations,” that differs from the traditional “memories-as-attractors” viewpoint. Memory recall from the memories-as-bifurcations viewpoint occurs when the spontaneous neural activity is changed to an appropriate output activity upon application of an input, known as a bifurcation in dynamical systems theory, wherein the input modifies the flow structure of the neural dynamics. Learning, then, is a process that helps create neural dynamical systems such that a target output pattern is generated as an attractor upon a given input. Based on this novel viewpoint, we introduce in this paper an associative memory model with a sequential learning process. Using a simple Hebbian-type learning, the model is able to memorize a large number of input/output mappings. The neural dynamics shaped through the learning exhibit different bifurcations to make the requested targets stable upon an increase in the input, and the neural activity in the absence of input shows chaotic dynamics with occasional approaches to the memorized target patterns. These results suggest that these dynamics facilitate the bifurcations to each target attractor upon application of the corresponding input, which thus increases the capacity for learning. This theoretical finding about the behavior of the spontaneous neural activity is consistent with recent experimental observations in which the neural activity without stimuli wanders among patterns evoked by previously applied signals. In addition, the neural networks shaped by learning properly reflect the correlations of input and target-output patterns in a similar manner to those designed in our previous study.

2PT224 Spontaneous and Evoked Neural Activities Shaped through A Sequential Learning Process(The 50th Annual Meeting of the Biophysical Society of Japan)

2PT224 Spontaneous and Evoked Neural Activities Shaped through A Sequential Learning Process(The 50th Annual Meeting of the Biophysical Society of Japan)

Adaptive FIR Neural Model for Centroid Learning in Self-Organizing Maps

In this paper, a training method for the formation of topology preserving maps is introduced. The proposed approach presents a sequential formulation of the self-organizing map (SOM), which is based on a new model of the neuron, or processing unit. Each neuron acts as a finite impulse response (FIR) system, and the coefficients of the filters are adaptively estimated during the sequential learning process, in order to minimize a distortion measure of the map. The proposed FIR-SOM model deals with static distributions and it computes an ordered set of centroids. Additionally, the FIR-SOM estimates the learning dynamic of each prototype using an adaptive FIR model. A noteworthy result is that the optimized coefficients of the FIR processes tend to represent a moving average filter, regardless of the underlying input distribution. The convergence of the resulting model is analyzed numerically and shows good properties with respect to the classic SOM and other unsupervised neural models. Finally, the optimal FIR coefficients are shown to be useful for visualizing the cluster densities.

Acquisition and memory of sequence order in young and adult chimpanzees (Pan troglodytes)

There is no research about age difference in the process of sequential learning in non-human primates. Is there any difference between young and adults in sequential learning process? Six chimpanzees (Pan troglodytes), 3 young and 3 adults, learned the Arabic numeral sequence 1 to 9 by touching the numerals on a touch-screen monitor in ascending order. Initially, the sequence always started with the numeral 1, i.e. 'start-fixed task'. Training began with the sequence 1-2, 1-2-3, and continued sequentially up to 1-2-3-4-5-6-7-8-9. Later, the subjects were introduced to sequences that started with a random numeral, but always ended with 9, i.e. 'end-fixed task'. Performance in the end-fixed task was worse relative to the familiar start-fixed task. After training with various sequences of adjacent numerals, the subjects were given a transfer test for the non-adjacent numerals. The results suggested that all chimpanzees indeed mastered sequential ordering, and although there was no fundamental difference in the acquisition process between the two age groups, there was a significant age difference in memory capacity. Based on their knowledge of sequential ordering, the subjects were then asked to perform a masking task in which once a subject touched the lowest numeral, the other numeral(s) turned to white squares. Performance of the masking task by young chimpanzees was better than that of adults in accuracy and degree of difficulty (number of numerals). Taken together, these data clearly demonstrate a similarity among subjects in the way chimpanzees acquire knowledge of sequential order regardless of age differences in sequential learning. Moreover, they reveal that once knowledge of sequential order is established, it can be a good index used to evaluate memory capacity in young and adult chimpanzees.

Optimization and stabilization of sequential learning in RBF network for nonlinear function approximation

This paper proposes a solution for inconsistency pruning of neurons within a sequential learning Radial Basis Function (RBF) Network. This paper adopts the concept that a specific RBF neuron which continuously exhibits low output in a sequence of training patterns does not justify the proposition that the neuron is insignificant to the whole function to be approximated. We establish additional criterions to provide protection from error in pruning RBF neurons within the hidden layer, which we prove is able to improve consistency and stability of neuron evolution. With such stability within the sequential learning process, we also show how the convergence speed of the network can be improved by reducing the number of consecutive observations required to prune a neuron in the hidden layer.