Fast Learning Research Articles

Development and Evaluation of An In-House Lumbar Puncture Simulator for First-Year Resident Lumbar Puncture Procedure Learning.

Lumbar puncture (LP) is a common invasive technique considered an essential learning milestone for anesthesiologists due to its application in spinalanesthesia. We aimed to develop an in-house LP simulator, test its effectiveness in learning the steps to perform anLP and analyze its impact on the first-year residents' self-confidence at our hospital. We used 3D printing and silicone casting to create an LP simulator based on a lumbar spinecomputed tomography (CT). We divided 12 first-year anesthesiology residents into control and experimental groups. The control group received traditional training, while the experimental group practiced with the simulator for three months. We used a procedure checklist anda Likert scalesurvey to evaluate their procedural knowledge and self-confidence at baseline, three, and six months. Eighteen months later, we evaluated their LP performance skills. Both groups showed a significant improvement in their knowledge scores over time.After three months, the experimental group had a higher median knowledge score (10 (10 - 10) median (min-max)) than the control group (9 (8 - 9.5) median (min-max)) (p = 0.03). While there were no apparent differences in median self-confidence scores between the groups at any time point, the experimental group had a significant increase in their self-confidence for performing an unassisted LP, with a median score of 1/5 (1 - 2.3) at baseline and 5/5 (4.8 - 5) after six months (p = 0.006). In contrast, the control group's self-confidence scores decreased from 4/5 (3 - 4) after three months to 3/5 (2 - 5) after six months. The evaluation of performance skills did not yield statistically significant results. Our study demonstrates that an in-house LP simulator is an effective and practical approach for first-year anesthesiology residents to learn the LP procedure. This approach could be particularly useful in settings with limited resources and a lack of sufficient patients to practice on, as it provides an opportunity for faster learning and increased self-confidence.

Virtual reality simulators for temporal bone dissection: overcoming limitations of previous models

Temporal bone dissection is a critical skill for otolaryngology trainees: however, it is challenging to master due to the complex anatomy and limited exposure to cadaveric specimens. The aim of this review was to develop and evaluate a novel virtual reality (VR) simulator for temporal bone dissection, addressing the limitations of previous simulators reported in the literature. A comprehensive literature search was conducted in the PubMed, Embase, and Cochrane Library databases from inception to September 2022. The search was limited to studies that evaluated the effectiveness of VR simulators for temporal bone dissection. The quality of the included studies was assessed using the Cochrane Risk of Bias Tool. The results showed that VR simulators can significantly improve temporal bone dissection skills, including anatomical knowledge, instrument handling, and surgical performance. Compared to traditional training methods, VR simulation was associated with faster learning curves, better retention of skills, and greater confidence among trainees. However, some limitations of current VR models were identified, including the lack of haptic feedback, limited realism, and short duration of practice. VR simulators are a valuable adjunct to traditional methods for temporal bone dissection training. The recently developed VR simulator addressed the limitations of previous simulators and demonstrated its potential to enhance the training of clinicians in temporal bone dissection. Future directions for research include further validation of the simulator and exploration of its potential for use in clinical settings.

A Two-Layer Self-Organizing Map with Vector Symbolic Architecture for Spatiotemporal Sequence Learning and Prediction.

We propose a new nature- and neuro-science-inspired algorithm for spatiotemporal learning and prediction based on sequential recall and vector symbolic architecture. A key novelty is the learning of spatial and temporal patterns as decoupled concepts where the temporal pattern sequences are constructed using the learned spatial patterns as an alphabet of elements. The decoupling, motivated by cognitive neuroscience research, provides the flexibility for fast and adaptive learning with dynamic changes to data and concept drift and as such is better suited for real-time learning and prediction. The algorithm further addresses several key computational requirements for predicting the next occurrences based on real-life spatiotemporal data, which have been found to be challenging with current state-of-the-art algorithms. Firstly, spatial and temporal patterns are detected using unsupervised learning from unlabeled data streams in changing environments; secondly, vector symbolic architecture (VSA) is used to manage variable-length sequences; and thirdly, hyper dimensional (HD) computing-based associative memory is used to facilitate the continuous prediction of the next occurrences in sequential patterns. The algorithm has been empirically evaluated using two benchmark and three time-series datasets to demonstrate its advantages compared to the state-of-the-art in spatiotemporal unsupervised sequence learning where the proposed ST-SOM algorithm is able to achieve 45% error reduction compared to HTM algorithm.

Faster nonconvex low-rank matrix learning for image low-level and high-level vision: A unified framework

This study introduces a unified approach to tackle challenges in both low-level and high-level vision tasks for image processing. The framework integrates faster nonconvex low-rank matrix computations and continuity techniques to yield efficient and high-quality results. In addressing real-world image complexities like noise, variations, and missing data, the framework exploits the intrinsic low-rank structure of the data and incorporates specific residual measurements. The optimization problem for low-rank matrix learning is effectively solved using the nonconvex Proximal Block Coordinate Descent (PBCD) algorithm, resulting in nearly unbiased estimators. Rigorous theoretical analysis ensures both local and global convergence. The PBCD algorithm updates blocks of variables iteratively with closed-form solutions, adeptly handling nonconvexity and promoting faster convergence. Notably, the framework incorporates the randomized singular value decomposition (RSVD) technique and introduces a continuous strategy for adaptive model parameter updates. These strategic choices reduce computational complexity while maintaining result quality. They offer fine-tuned control over the desired rank of the learned matrix and enhance robustness in a straightforward manner. Furthermore, the versatility of the proposed nonconvex PBCD algorithm extends to handling problems with multiple variables, as supported by theoretical analysis. Experimental evaluations, spanning various image low-level and high-level vision tasks such as inpainting, classification, and clustering, validate the effectiveness and efficiency of our framework across diverse databases. The source code is available at https://github.com/ZhangHengMin/FNPBCD_LR.In a nutshell, our framework provides a unified solution to tackle both low-level and high-level vision tasks in images. By combining fast nonconvex low-rank matrix learning with adaptive parameter updates, we achieve efficient computation, yielding high-quality results that demonstrate robustness against various types of noise. The evaluations further endorse the reliability and applicability of our proposed framework.

Diabetes detection based on Fast Learning Network model

In the last decades, machine learning algorithms have witnessed a high significance in healthcare applications in terms of detecting various diseases. For instance, diabetic disease is considered one of the major health problems around the world. However, there is a need to deeply study a machine learning algorithm in the detection of diabetic disease. Therefore, this study presents a Fast Learning Network (FLN) algorithm in the detection of diabetic disease based on different numbers of hidden nodes. In this work, the Pima Indians Diabetes Database (PIDD) is used for training and testing the proposed FLN algorithm. Furthermore, the performance of the proposed model has been assessed in terms of several evaluation measurements such as accuracy, precision, recall, F-Measure, G-Mean, MCC, and specificity. The experimental results show that the highest achieved accuracy, recall, F-Measure, G-Mean, and MCC were 82.17%, 80.95%, 71.33%, 71.84%, and 59.54%, respectively. Meanwhile, the highest obtained results for precision and specificity were 67.50% and 83.12%, respectively. In addition, the performance of the proposed model has outperformed its comparative in terms of detection accuracy.

Impact of enriched environment on motor performance and learning in mice

Neuroscience heavily relies on animal welfare in laboratory rodents as it can significantly affect brain development, cognitive function and memory formation. Unfortunately, laboratory animals are often raised in artificial environments devoid of physical and social stimuli, potentially leading to biased outcomes in behavioural assays. To assess this effect, we examined the impact of social and physical cage enrichment on various forms of motor coordination. Our findings indicate that while enriched-housed animals did not exhibit faster learning in eyeblink conditioning, the peak timing of their conditioned responses was slightly, but significantly, improved. Additionally, enriched-housed animals outperformed animals that were housed in standard conditions in the accelerating rotarod and ErasmusLadder test. In contrast, we found no significant effect of enrichment on the balance beam and grip strength test. Overall, our data suggest that an enriched environment can improve motor performance and motor learning under challenging and/or novel circumstances, possibly reflecting an altered state of anxiety.

Personalized Federated Learning with Adaptive Feature Extraction and Category Prediction in Non-IID Datasets

Federated learning trains a neural network model using the client’s data to maintain the benefits of centralized model training while maintaining their privacy. However, if the client data are not independently and identically distributed (non-IID) because of different environments, the accuracy of the model may suffer from client drift during training owing to discrepancies in each client’s data. This study proposes a personalized federated learning algorithm based on the concept of multitask learning to divide each client model into two layers: a feature extraction layer and a category prediction layer. The feature extraction layer maps the input data to a low-dimensional feature vector space. Furthermore, the parameters of the neural network are aggregated with those of other clients using an adaptive method. The category prediction layer maps low-dimensional feature vectors to the label sample space, with its parameters remaining unaffected by other clients to maintain client uniqueness. The proposed personalized federated learning method produces faster learning model convergence rates and higher accuracy rates for the non-IID datasets in our experiments.

Hierarchical matrix factorization for interpretable collaborative filtering

Matrix factorization (MF) is a simple collaborative filtering technique that achieves superior recommendation accuracy by decomposing the user–item interaction matrix into user and item latent matrices. Because the model typically learns each interaction independently, it may overlook the underlying shared dependencies between users and items, resulting in less stable and interpretable recommendations. Based on these insights, we propose “Hierarchical Matrix Factorization” (HMF), which incorporates clustering concepts to capture the hierarchy, where leaf nodes and other nodes correspond to users/items and clusters, respectively. Central to our approach, called hierarchical embeddings, is the additional decomposition of the latent matrices (embeddings) into probabilistic connection matrices, which link the hierarchy, and a root cluster latent matrix. The embeddings are differentiable, allowing simultaneous learning of interactions and clustering using a single gradient descent method. Furthermore, the obtained cluster-specific interactions naturally summarize user–item interactions and provide interpretability. Experimental results on ratings and ranking predictions show that HMF outperforms existing MF methods, in particular achieving a 1.37 point improvement in RMSE for sparse interactions. Additionally, it was confirmed that the clustering integration of HMF has the potential for faster learning convergence and mitigation of overfitting compared to MF, and also provides interpretability through a cluster-centered case study.

Visual perceptual learning of feature conjunctions leverages non-linear mixed selectivity

Visual objects are often defined by multiple features. Therefore, learning novel objects entails learning feature conjunctions. Visual cortex is organized into distinct anatomical compartments, each of which is devoted to processing a single feature. A prime example are neurons purely selective to color and orientation, respectively. However, neurons that jointly encode multiple features (mixed selectivity) also exist across the brain and play critical roles in a multitude of tasks. Here, we sought to uncover the optimal policy that our brain adapts to achieve conjunction learning using these available resources. 59 human subjects practiced orientation-color conjunction learning in four psychophysical experiments designed to nudge the visual system towards using one or the other resource. We find that conjunction learning is possible by linear mixing of pure color and orientation information, but that more and faster learning takes place when both pure and mixed selectivity representations are involved. We also find that learning with mixed selectivity confers advantages in performing an untrained “exclusive or” (XOR) task several months after learning the original conjunction task. This study sheds light on possible mechanisms underlying conjunction learning and highlights the importance of learning by mixed selectivity.

Driving Organizational Effectiveness: Implementing Safe Agile Framework for Team Alignment in Large Organisations

Implementing the Scaled Agile Framework (SAFe) in large organizations is crucial for achieving organizational effectiveness, especially as they expand globally. The SAFe framework, initially designed for software development, has evolved to cater to diverse industries like health, manufacturing, engineering, and financial services. The study elaborates the significance of implementing SAFe agile framework for team alignment in large organisations. SAFe operates at four levels: Team, Program, Large Solution, and Portfolio. Efficient collaboration, communication, and coordination, the three Cs of SAFe Agile, are essential for synergy across these levels. To practically implement SAFe and achieve team alignment, organizations should follow ten SAFe principles. These principles include taking an economic view, applying systems thinking, assuming variability, building incrementally with fast learning cycles, basing milestones on objective evaluation, ensuring uninterrupted value flow, applying cadence, unlocking intrinsic motivation, embracing decentralized decision-making, and organizing around value. Balancing bottom-up and top-down approaches through a middle-out method ensures effective SAFe implementation. This approach combines the strengths of methods, fostering cross-functional collaboration and effective communication channels from teams to the portfolio level. In conclusion, large-scale businesses may enhance organizational success through the transformational process of applying the SAFe framework for team alignment, which improves alignment, cooperation, efficiency, flexibility, and continuous improvement.

FORECASTING THE NUMBER OF AIRPLANE PASSENGERS USING HOLT WINTER'S EXPONENTIAL SMOOTHING METHOD AND EXTREME LEARNING MACHINE METHOD

Airplanes provide comfort and speed for their users, especially for those who have limited time. The number of passengers has continued to increase in the last few months at Ahmad Yani International Airport, so a forecast is needed in making decisions to predict the number of passengers in order to maximize existing performance. The data used is secondary data on the number of airplane passengers at Ahmad Yani International Airport from 2012 to 2022 obtained from PT Angkasa Pura 1 (Persero). The Holt Winters Exponential Smoothing method is used because it aligns with the data pattern that includes trends and seasonality in the research, and it has a low level of accuracy. In this study also used the Extreme Learning Machine (ELM) method, apart from being a relatively new method, it has a fast learning speed and has low accuracy. This study aims to predict the number of airplane passengers at Ahmad Yani International Airport in Semarang using the Holt Winters Exponential Smoothing and ELM methods. The results of the analysis show that the MAPE value in the Holt Winters Exponential Smoothing method is 8,18% and in the ELM method using 12 input neurons and 43 neurons in the hidden layer, a MAPE of 6,04% is obtained. so that the ELM method is the right method for predicting the number of airplane passengers at Ahmad Yani International Airport in Semarang.

플레이 수준 조절이 가능한 강화학습 기반 카드형 대전 게임 에이전트

Game agents which are behavioral agent for game playing are a crucial component of game satisfaction. However it takes a lot of time and effort to create game agents for various game levels, environments, and players. In addition, when the game environment changes such as adding contents or updating characters, new game agents need to be developed and the development difficulty gradually increases. And it is important to have a game agent that can be customized for different levels of players. This is because a game agent that can play games of various levels is more useful and can increase the satisfaction of more players than a high-level game agent. In this paper, we propose a method for learning and controlling the level of play of game agents that can be rapidly developed and fine-tuned for various game environments and changes. At this time, reinforcement learning applies a policy-based distributed reinforcement learning method IMPALA for flexible processing and fast learning of various behavioral structures. Once reinforcement learning is complete, we choose actions by sampling based on Softmax-Temperature method. From this result, we show that the game agent's play level decreases as the Temperature value increases. This shows that it is possible to easily control the play level.

Improving extreme learning machine model using deep learning feature extraction and grey wolf optimizer: Application to image classification

The Extreme Learning Machine (ELM) is a highly efficient model for real-time network retraining due to its fast learning speed, unlike traditional machine learning methods. However, the performance of ELM can be negatively impacted by the random initialization of weights and biases. Moreover, poor input feature quality can further degrade performance, particularly with complex visual data. To overcome these issues, this paper proposes optimizing the input features as well as the initial weights and biases. We combine both Convolutional Neural Network (CNN) and Convolutional AutoEncoder (CAE) extracted features to optimize the quality of the input features. And we use our hybrid Grey Wolf Optimizer-Multi-Verse Optimizer (GWO-MVO) metaheuristic for initializing weights and biases by applying four fitness functions based on: the norm of the output weights, the error rate on the training set, and the error rate on the validation set. Our method is evaluated on image classification tasks using two benchmark datasets: CIFAR-10 and CIFAR-100. Since image quality may vary in real-world applications, we trained and tested our models on the dataset’s original and noisy versions. The results demonstrate that our method provides a robust and efficient alternative for image classification tasks, offering improved accuracy and reduced overfitting.

Neurons in the monkey frontopolar cortex encode learning stage and goal during a fast learning task.

The frontopolar cortex (FPC) is, to date, one of the least understood regions of the prefrontal cortex. The current understanding of its function suggests that it plays a role in the control of exploratory behaviors by coordinating the activities of other prefrontal cortex areas involved in decision-making and exploiting actions based on their outcomes. Based on this hypothesis, FPC would drive fast-learning processes through a valuation of the different alternatives. In our study, we used a modified version of a well-known paradigm, the object-in-place (OIP) task, to test this hypothesis in electrophysiology. This paradigm is designed to maximize learning, enabling monkeys to learn in one trial, which is an ability specifically impaired after a lesion of the FPC. We showed that FPC neurons presented an extremely specific pattern of activity by representing the learning stage, exploration versus exploitation, and the goal of the action. However, our results do not support the hypothesis that neurons in the frontal pole compute an evaluation of different alternatives. Indeed, the position of the chosen target was strongly encoded at its acquisition, but the position of the unchosen target was not. Once learned, this representation was also found at the problem presentation, suggesting a monitoring activity of the synthetic goal preceding its acquisition. Our results highlight important features of FPC neurons in fast-learning processes without confirming their role in the disengagement of cognitive control from the current goals.

Lightweight and high-precision materials property prediction using pre-trained Graph Neural Networks and its application to small dataset

Abstract Large data sets are essential for building deep learning models. However, generating large datasets with higher theoretical levels and larger computational models remains difficult due to the high cost of first-principles calculation. Here, we propose a lightweight and highly accurate machine learning approach using pre-trained Graph Neural Networks (GNNs) for industrially important but difficult to scale models. The proposed method was applied to a small dataset of graphene surface systems containing surface defects, and achieved comparable accuracy with six orders of magnitude faster learning than when the GNN was trained from scratch.

Real-time sign language recognition based on YOLO algorithm

This study focuses on real-time hand gesture recognition in the Turkish sign language detection system. YOLOv4-CSP based on convolutional neural network (CNN), a state-of-the-art object detection algorithm, is used to provide real-time and high-performance detection. The YOLOv4-CSP algorithm is created by adding CSPNet to the neck of the original YOLOv4 to improve network performance. A new object detection model has been proposed by optimizing the YOLOv4-CSP algorithm in order to provide more efficient detection in Turkish sign language. The model uses CSPNet throughout the network to increase the learning ability of the network. However, Proposed YOLOv4-CSP has a learning model with Mish activation function, complete intersection of union (CIoU) loss function and transformer block added. The Proposed YOLOv4-CSP algorithm has faster learning with transfer learning than previous versions. This allows the proposed YOLOv4-CSP algorithm to perform a faster restriction and recognition of static hand signals simultaneously. To evaluate the speed and detection performance of the proposed YOLOv4-CSP model, it is compared with previous YOLO series, which offers real-time detection, as well. YOLOv3, YOLOv3-SPP, YOLOv4-CSP and proposed YOLOv4-CSP models are trained with a labeled dataset consisting of numbers in Turkish Sign language, and their performances on the hand signals recognitions are compared. With the proposed method, 98.95% precision, 98.15% recall, 98.55 F1 score and 99.49% mAP results are obtained in 9.8 ms. The proposed method for detecting numbers in Turkish sign language outperforms other algorithms with both real-time performance and accurate hand sign prediction, regardless of background.

Features of adaptive training algorithms for improved complex skill acquisition

Training complex skills is typically accomplished by means of a trainer or mediator who tailors instruction to the individual trainee. However, facilitated training is costly and labor intensive, and the use of a mediator is infeasible in remote or extreme environments. Imparting complex skills in applications like long-duration human spaceflight, military field operations, or remote medicine may require automated training algorithms. Virtual reality (VR) is an effective, easily programmable, immersive training medium that has been used widely across fields. However, there remain open questions in the search for the most effective algorithms for guiding automated training progression. This study investigates the effects of responsiveness, personalization, and subtask independence on the efficacy of automated training algorithms in VR for training complex, operationally relevant tasks. Thirty-two subjects (16M/16F, 18–54 years) were trained to pilot and land a spacecraft on Mars within a VR simulation using four different automated training algorithms. Performance was assessed in a physical cockpit mock-up. We found that personalization results in faster skill acquisition on average when compared with a standardized progression built for a median subject (p = 0.0050). The standardized progression may be preferable when consistent results are desired across all subjects. Independence of the difficulty adjustments between subtasks may lead to increased skill acquisition, while lockstep in the progression of each subtask increases self-reported flow experience (p = 0.01), fluency (p = 0.02), and absorption (p = 0.01) on the Flow Short Scale. Data visualization suggests that highly responsive algorithms may lead to faster learning progressions and higher skill acquisition for some subjects. Improving transfer of skills from training to testing may require either high responsiveness or a standardized training progression. Optimizing the design of automated, individually adaptive algorithms around the training needs of a group may be useful to increase skill acquisition for complex operational tasks.

The learning curve for using intestinal ultrasonography

Background and aimsIntestinal ultrasonography (IUS) is challenging to learn. This prospective study examined how the accuracy of IUS increases with operator experience (“learning curve”) and if prior abdominal ultrasound experience facilitates the learning process. MethodsThe study included two trainees with limited abdominal ultrasound experience (< 50 exams) and two with extensive experience (> 500 exams). Each trainee performed 99 examinations and reported four IUS findings. An expert sonographer repeated the exam, and concordance (k) between the expert and trainees was assessed in three consecutive testing periods of 33 exams each. ResultsA progressive improvement in concordance was observed for all IUS findings from Period 1 to Period 3, overall and for both groups of trainees, although those with experience in abdominal ultrasound had faster learning curves. The minimum number of examinations required to achieve concordance with the expert operator for detecting increased bowel wall thickness was 84 and detecting bowel dilatation was 79. However, a minimum of 97 examinations was necessary to achieve concordance for detecting intra-abdominal complications, considered an advanced IUS competence. ConclusionBasic competence in IUS can be acquired with relatively few examinations, while advanced competence requires more extensive training, particularly for gastroenterologists without abdominal ultrasound experience.

Extreme Learning Machine Classification Method based on Twin Strategy Salp Swarm Algorithm

Extreme Learning Machines (ELM) are a type of Single Hidden Layer Feedforward Neural Network (SLFN). In recent years, they have attracted attention for their powerful approximation ability and fast learning speed. Compared with traditional neural network algorithms, ELM have the advantages of simple structure, fast learning speed, and good generalization performance. However, since the input weights and biases of ELM are randomly generated, there may be some suboptimal or unnecessary input weights and biases. In addition, ELM may require more hidden nodes, which may slow down its response to unknown test data. To address these issues, a twin strategy Salp Swarm Algorithm (TSSA) is proposed to increase the population diversity to improve the convergence speed of the algorithm through chaotic initialization, while shock inertia weights and learning paradigms are introduced into the follower position updating to enhance the stochasticity of the particles. Classification tests are conducted on six datasets using the proposed TSSA, and the experimental results show that the proposed TSSA-ELM has higher accuracy and better generalization performance than some existing ELM variants.

Fast random opposition-based learning Aquila optimization algorithm

Meta-heuristic algorithms are usually employed to address a variety of challenging optimization problems. In recent years, there has been a continuous effort to develop new and efficient meta-heuristic algorithms. The Aquila Optimization (AO) algorithm is a newly established swarm-based method that mimics the hunting strategy of Aquila birds in nature. However, in complex optimization problems, the AO has shown a sluggish convergence rate and gets stuck in the local optimal region throughout the optimization process. To overcome this problem, in this study, a new mechanism named Fast Random Opposition-Based Learning (FROBL) is combined with the AO algorithm to improve the optimization process. The proposed approach is called the FROBLAO algorithm. To validate the performance of the FROBLAO algorithm, the CEC 2005, CEC 2019, and CEC 2020 test functions, along with six real-life engineering optimization problems, are tested. Moreover, statistical analyses such as the Wilcoxon rank-sum test, the t-test, and the Friedman test are performed to analyze the significant difference between the proposed algorithm FROBLAO and other algorithms. The results demonstrate that FROBLAO achieved outstanding performance and effectiveness in solving an extensive variety of optimization problems.