Size Of Training Data Research Articles

Measurement-Based Neural Network Technique for Modeling the Low-Frequency Electric Field Radiated Behavior of Satellite Units

In this work, a machine learning technique employing a measurement-driven artificial neural network (ANN) architecture is proposed as a solution to the precise determination of the position and the moment of equivalent electric dipoles for unit characterization. These dipoles are used to match the generated electric field from various sources inside the spacecraft, during space exploration missions. Various methodologies for unit characterization have been proposed in the literature, the most common being the heuristic approaches, least squares variants, method of auxiliary sources, etc. Contrary to the previous time-consuming post-process methodologies, the proposed electric dipole neural network (EDMnet) can offer a real-time characterization of the measured unit (Device Under Test) after a proper training stage, especially as a fast pre-compliance method. The network uses the electric field vector, measured at 14 discrete locations, as input and reports the position and moment of the electric dipole that best matches the measured fields. In this work, various ANN architectures are tested and compared in order to select the optimal EDMnet parameters for accurate source identification. It is shown that the size of the artificial training data affects the performance of the network. The proposed EDMnet can provide accuracy in mm-scale, with respect to dipole positioning, greater than 99% in dipole moment prediction.

Soot temperature and volume fraction field predictions via line-of-sight soot integral radiation equation informed neural networks in laminar sooting flames

This paper originally proposes a physics informed neural networks (PINNs) model for the simultaneous prediction of soot temperature and volume fraction fields in laminar flames from experimental soot integral radiation. Contrasted with the previous data-driven models, the PINNs model incorporates the line-of-sight soot radiation integral equation into the model architecture. Doing so, the superiority of the physics informed neural networks model is displayed in terms of prediction accuracy under limited training data size and training efficiency. Due to the significant reduction of training experimental data dependence, such gray-box physics informed methodology sheds light on practical combustion devices combustion monitoring, i.e., limited optical window endoscope engines.

Bilateral symmetry-based augmentation method for improved tooth segmentation in panoramic X-rays

Panoramic X-rays are crucial in dental radiology, providing detailed images that are essential for diagnosing and planning treatment for various oral conditions. The advent of automated methods that learn from annotated data promises to significantly aid clinical experts in making accurate diagnoses. However, these methods often require large amounts of annotated data, making the generation of high-quality annotations for panoramic X-rays both challenging and time-consuming. This paper introduces a novel bilateral symmetry-based augmentation method specifically designed to enhance tooth segmentation in panoramic X-rays. By exploiting the inherent bilateral symmetry of these images, our proposed method systematically generates augmented data, leading to substantial improvements in the performance of tooth segmentation models. By increasing the training data size fourfold, our approach proportionately reduces the effort required to manually annotate extensive datasets. These findings highlight the potential of leveraging the symmetrical properties of medical images to enhance model performance and accuracy in dental radiology. The effectiveness of the proposed method is evaluated on three widely adopted deep learning models: U-Net, SE U-Net, and TransUNet. Significant improvements in segmentation accuracy are observed with the proposed augmentation method across all models. For example, the average Dice Similarity Coefficient (DSC) increases by over 8%, reaching 76.7% for TransUNet. Further, comparisons with existing augmentation methods, including rigid transform-based and elastic grid-based techniques, show that the proposed method consistently outperforms them with additional improvements up to 5% in terms of average DSC, with the exact improvement varying depending on the model and training dataset size. We have made the data augmentation codes and tools developed based on our method available at https://github.com/wathoresanket/bilateralsymmetrybasedaugmentation.

Predicting individual patient and hospital-level discharge using machine learning

BackgroundAccurately predicting hospital discharge events could help improve patient flow and the efficiency of healthcare delivery. However, using machine learning and diverse electronic health record (EHR) data for this task remains incompletely explored.MethodsWe used EHR data from February-2017 to January-2020 from Oxfordshire, UK to predict hospital discharges in the next 24 h. We fitted separate extreme gradient boosting models for elective and emergency admissions, trained on the first two years of data and tested on the final year of data. We examined individual-level and hospital-level model performance and evaluated the impact of training data size and recency, prediction time, and performance in subgroups.ResultsOur models achieve AUROCs of 0.87 and 0.86, AUPRCs of 0.66 and 0.64, and F1 scores of 0.61 and 0.59 for elective and emergency admissions, respectively. These models outperform a logistic regression model using the same features and are substantially better than a baseline logistic regression model with more limited features. Notably, the relative performance increase from adding additional features is greater than the increase from using a sophisticated model. Aggregating individual probabilities, daily total discharge estimates are accurate with mean absolute errors of 8.9% (elective) and 4.9% (emergency). The most informative predictors include antibiotic prescriptions, medications, and hospital capacity factors. Performance remains robust across patient subgroups and different training strategies, but is lower in patients with longer admissions and those who died in hospital.ConclusionsOur findings highlight the potential of machine learning in optimising hospital patient flow and facilitating patient care and recovery.

Classification of images derived from submarine fibre optic sensing: detecting broadband seismic activity from hydroacoustic signals

Abstract Distributed acoustic sensing (DAS) is an optoelectronic technology that utilises fibre optic cables to detect disturbances caused by seismic waves. Using DAS, seismologists can monitor geophysical phenomena at high spatial and temporal resolutions over long distances in inhospitable environments. Field experiments using DAS, are typically associated with large volumes of observations, requiring algorithms for efficient processing and monitoring capabilities. In this study, we present a supervised classifier trained to recognise seismic activity from other sources of hydroacoustic energy. Our classifier is based on a 2D convolutional neural network architecture. The 55km-long ocean-bottom fibre optic cable, located off Cape Muroto in south-west of Japan, was interrogated using DAS. Data were collected during two different monitoring time-periods. Optimisation of the model’s hyperparameters using Gaussian Processes Regression was necessary to prevent issues associated with small sizes of training data. Using a test set of 100 annotated images, we have shown that the top performing model is around $92\%$ accurate in classifying geophysical data from other sources of hydroacoustic energy and ambient noise.

Learning properties of quantum states without the IID assumption

We develop a framework for learning properties of quantum states beyond the assumption of independent and identically distributed (i.i.d.) input states. We prove that, given any learning problem (under reasonable assumptions), an algorithm designed for i.i.d. input states can be adapted to handle input states of any nature, albeit at the expense of a polynomial increase in training data size (aka sample complexity). Importantly, this polynomial increase in sample complexity can be substantially improved to polylogarithmic if the learning algorithm in question only requires non-adaptive, single-copy measurements. Among other applications, this allows us to generalize the classical shadow framework to the non-i.i.d. setting while only incurring a comparatively small loss in sample efficiency. We leverage permutation invariance and randomized single-copy measurements to derive a new quantum de Finetti theorem that mainly addresses measurement outcome statistics and, in turn, scales much more favorably in Hilbert space dimension.

Leveraging Unsupervised Task Adaptation and Semi‐Supervised Learning With Semantic‐Enriched Representations for Online Sexism Detection

ABSTRACTOver the past decade, the proliferation of hateful and sexist content targeting women on social media has become a concerning issue, adversely affecting women's lives and freedom of expression. Previous efforts to detect online sexism have utilized monolingual ensemble transformers combined with data augmentation techniques that incorporate related‐domain data, such as hate speech. However, these approaches often struggle to capture the full diversity and complexity of sexism due to limitations in the size and quality of training data. In this study, we introduce a novel sexism detection system that employs in‐domain unlabeled data through unsupervised task‐adaptation techniques and semi‐supervised learning, using an efficient single multilingual transformer model. Additionally, we incorporate a Sentence‐BERT layer to enhance our system with semantically meaningful sentence embeddings. Our proposed system outperforms existing state‐of‐the‐art methods across all tasks and datasets, demonstrating its effectiveness in detecting and addressing sexism in social media text. These results underscore the potential of our approach, providing a foundation for further research and practical applications.

Fine-grained Automatic Augmentation for handwritten character recognition

With the advancement of deep learning-based character recognition models, the training data size has become a crucial factor in improving the performance of handwritten text recognition. For languages with low-resource handwriting samples, data augmentation methods can effectively scale up the data size and improve the performance of handwriting recognition models. However, existing data augmentation methods for handwritten text face two limitations: (1) Methods based on global spatial transformations typically augment the training data by transforming each word sample as a whole but ignore the potential to generate fine-grained transformation from local word areas, limiting the diversity of the generated samples; (2) It is challenging to adaptively choose a reasonable augmentation parameter when applying these methods to different language datasets. To address these issues, this paper proposes Fine-grained Automatic Augmentation (FgAA) for handwritten character recognition. Specifically, FgAA views each word sample as composed of multiple strokes and achieves data augmentation by performing fine-grained transformations on the strokes. Each word is automatically segmented into various strokes, and each stroke is fitted with a Bézier curve. On such a basis, we define the augmentation policy related to the fine-grained transformation and use Bayesian optimization to select the optimal augmentation policy automatically, thereby achieving the automatic augmentation of handwriting samples. Experiments on seven handwriting datasets of different languages demonstrate that FgAA achieves the best augmentation effect for handwritten character recognition. Our code is available at https://github.com/IMU-MachineLearningSXD/Fine-grained-Automatic-Augmentation

Transfer-learning enabled micro-expression recognition using dense connections and mixed attention

Micro-expression recognition (MER) is a challenging computer vision problem, where the limited amount of available training data and insufficient intensity of the facial expressions are among the main issues adversely affecting the performance of existing recognition models. To address these challenges, this paper explores a transfer–learning enabled MER model using a densely connected feature extraction module with mixed attention. Unlike previous works that utilize transfer learning to facilitate MER and extract local facial-expression information, our model relies on pretraining with three diverse macro-expression datasets and, as a result, can: (i) overcome the problem of insufficient sample size and limited training data availability, (ii) leverage (related) domain-specific information from multiple datasets with diverse characteristics, and (iii) improve the model adaptability to complex scenes. Furthermore, to enhance the intensity of the micro-expressions and improve the discriminability of the extracted features, the Euler video magnification (EVM) method is adopted in the preprocessing stage and then used jointly with a densely connected feature extraction module and a mixed attention mechanism to derive expressive feature representations for the classification procedure. The proposed feature extraction mechanism not only guarantees the integrity of the extracted features but also efficiently captures local texture cues by aggregating the most salient information from the generated feature maps, which is key for the MER task. The experimental results on multiple datasets demonstrate the robustness and effectiveness of our model compared to the state-of-the-art.

Linking datasets on organizations using half a billion open-collaborated records

Abstract Scholars studying organizations often work with multiple datasets lacking shared identifiers or covariates. In such situations, researchers usually use approximate string (“fuzzy”) matching methods to combine datasets. String matching, although useful, faces fundamental challenges. Even where two strings appear similar to humans, fuzzy matching often struggles because it fails to adapt to the informativeness of the character combinations. In response, a number of machine learning methods have been developed to refine string matching. Yet, the effectiveness of these methods is limited by the size and diversity of training data. This paper introduces data from a prominent employment networking site (LinkedIn) as a massive training corpus to address these limitations. By leveraging information from the LinkedIn corpus regarding organizational name-to-name links, we incorporate trillions of name pair examples into various methods to enhance existing matching benchmarks and performance by explicitly maximizing match probabilities. We also show how relationships between organization names can be modeled using a network representation of the LinkedIn data. In illustrative merging tasks involving lobbying firms, we document improvements when using the LinkedIn corpus in matching calibration and make all data and methods open source.

Sensitivity Analysis of Traffic Sign Recognition to Image Alteration and Training Data Size

Sensitivity Analysis of Traffic Sign Recognition to Image Alteration and Training Data Size

An Adaptively Weighted Averaging Method for Regional Time Series Extraction of fMRI-Based Brain Decoding.

Brain decoding that classifies cognitive states using the functional fluctuations of the brain can provide insightful information for understanding the brain mechanisms of cognitive functions. Among the common procedures of decoding the brain cognitive states with functional magnetic resonance imaging (fMRI), extracting the time series of each brain region after brain parcellation traditionally averages across the voxels within a brain region. This neglects the spatial information among the voxels and the requirement of extracting information for the downstream tasks. In this study, we propose to use a fully connected neural network that is jointly trained with the brain decoder to perform an adaptively weighted average across the voxels within each brain region. We perform extensive evaluations by cognitive state decoding, manifold learning, and interpretability analysis on the Human Connectome Project (HCP) dataset. The performance comparison of the cognitive state decoding presents an accuracy increase of up to 5% and stable accuracy improvement under different time window sizes, resampling sizes, and training data sizes. The results of manifold learning show that our method presents a considerable separability among cognitive states and basically excludes subject-specific information. The interpretability analysis shows that our method can identify reasonable brain regions corresponding to each cognitive state. Our study would aid the improvement of the basic pipeline of fMRI processing.

Fintech model performance analysis in credit loans: evidence from China city commercial banks

ABSTRACT The analysis of bank loan performance has consistently been a central issue in banking research. With the impact of the COVID-19 pandemic, we have witnessed a significant increase in online loans. Our study employs fintech models to develop a big data credit analysis system, evaluating model efficiency and the effect of ‘data organization’ on predictive performance for city commercial bank loans. Our results show that training data size, timeliness, and inter-bank ‘data sharing’ significantly affect model predictions. We also find that traditional econometric models (Logit and Probit) have the weakest predictive ability. However, ‘most advanced’ deep learning models perform worse than machine learning models with lower algorithmic complexity, such as Random Forest. Besides, ‘sharing training data’ generally increases model efficiency, but there are exceptions for larger city commercial banks. This paper provides ample empirical evidence for the application of the fintech model in bank loan credit. It highlights several characteristics of training data, such as size and timeliness, which greatly affect model efficiencies. We also provide policy recommendations for fintech transition and collaboration among municipal banks.

An exponential reduction in training data sizes for machine learning derived entanglement witnesses

Abstract We propose a support vector machine (SVM) based approach for generating an entanglement witness that requires exponentially less training data than previously proposed methods. SVMs generate hyperplanes represented by a weighted sum of expectation values of local observables whose coefficients are optimized to sum to a positive number for all separable states and a negative number for as many entangled states as possible near a specific target state. Previous SVM-based approaches for entanglement witness generation used large amounts of randomly generated separable states to perform training, a task with considerable computational overhead. Here, we propose a method for orienting the witness hyperplane using only the significantly smaller set of states consisting of the eigenstates of the generalized Pauli matrices and a set of entangled states near the target entangled states. With the orientation of the witness hyperplane set by the SVM, we tune the plane's placement using a differential program that ensures perfect classification accuracy on a limited test set as well as maximal noise tolerance. For \(N\) qubits, the SVM portion of this approach requires only \(O(6^N)\) training states, whereas an existing method needs \(O(2^{4^N})\). We use this method to construct witnesses of 4 and 5 qubit GHZ states with coefficients agreeing with stabilizer formalism witnesses to within 3.7 percent and 1 percent, respectively. We also use the same training states to generate novel 4 and 5 qubit W state witnesses. Finally, we computationally verify these witnesses on small test sets and propose methods for further verification.

Mussel culture monitoring with semi-supervised machine learning on multibeam echosounder data using label spreading

High diversity seabed habitats, such as shellfish aggregations, play a significant role in marine ecosystem sustainability but are susceptible to bottom disturbance induced by anthropogenic activities. Regular monitoring of these habitats with effective mapping methods is therefore essential. Multibeam echosounder (MBES) has been widely used in recent decades for seabed characterization due to its non-destructive manner and extensive spatial coverage compared to traditional methods like bottom sampling. Nevertheless, bottom sampling remains essential to link ground truth with acoustic seabed classification. Using seabed samples and MBES measurements, machine learning techniques are commonly employed to model their relationships and generate classification maps of an extended seabed. However, limited ground truth data, resulting from constraints in regulations, budget, or time, may impede the development of robust machine learning models. To address this challenge, we applied a semi-supervised machine learning method to classify seabed sediments of a blue mussel (Mytilus edulis) cultivation area in the Oosterschelde, the Netherlands. We utilized nine boxcore samples to generate pseudo-labels on MBES data. These pseudo-labels enlarged the training data size, facilitated the training of three comprehensive machine learning algorithms (Gradient Boosting, Random Forest, and Support Vector Machine), and helped to classify the study site into mussel and non-mussel areas. We found the geomorphological and backscatter-related features to be complementary for mussel culture detection. Our classification results were demonstrated effective through expert knowledge of this cultivation area and brought insights for future research on natural mussel habitats.

Leakage Identification of Underground Structures Using Classification Deep Neural Networks and Transfer Learning.

Water leakage defects often occur in underground structures, leading to accelerated structural aging and threatening structural safety. Leakage identification can detect early diseases of underground structures and provide important guidance for reinforcement and maintenance. Deep learning-based computer vision methods have been rapidly developed and widely used in many fields. However, establishing a deep learning model for underground structure leakage identification usually requires a lot of training data on leakage defects, which is very expensive. To overcome the data shortage, a deep neural network method for leakage identification is developed based on transfer learning in this paper. For comparison, four famous classification models, including VGG16, AlexNet, SqueezeNet, and ResNet18, are constructed. To train the classification models, a transfer learning strategy is developed, and a dataset of underground structure leakage is created. Finally, the classification performance on the leakage dataset of different deep learning models is comparatively studied under different sizes of training data. The results showed that the VGG16, AlexNet, and SqueezeNet models with transfer learning can overall provide higher and more stable classification performance on the leakage dataset than those without transfer learning. The ResNet18 model with transfer learning can overall provide a similar value of classification performance on the leakage dataset than that without transfer learning, but its classification performance is more stable than that without transfer learning. In addition, the SqueezeNet model obtains an overall higher and more stable performance than the comparative models on the leakage dataset for all classification metrics.

Automatic phishing website detection and prevention model using transformer deep belief network

In the digitally connected world cybersecurity is paramount, phishing where attackers pose as trusted entities to steal sensitive data, looms large. The proliferation of phishing attacks on the internet poses a substantial threat to individuals and organizations, compromising sensitive information and causing financial and reputational damage. This study's goal is to establish an automated system for the early detection and prevention of phishing websites, thereby enhancing online security and protecting users from cyber threats. This research initially employs One Hot Encoding (OHE) mechanism-based pre-processing mechanism that converts every URL string into a numerical vector with a particular dimension. This study utilizes two feature selection techniques which are transfer learning-based feature extraction using DarkNet19 and Variational Autoencoder (VAE) to select the value of the most important feature. The robust security mechanisms are presented to prevent phishing attacks and safeguard personal information on websites. List-based deep learning-based systems to prevent and detect phishing URLs more efficiently. The study proposes a transformer-based Deep Belief Network (TB-DBN), a veritable pre-trained deep transformer network model for phishing behaviour detection. A cross-validation technique with grid search hyper-parameter tuning based on the Intelligence Binary Bat Algorithm (IBBA) was designed using the proposed hybrid model. Predictions were made to classify the phishing URLs using a probabilistic estimation guided boosting classifier model and evaluate their performance in terms of accuracy, precision, recall, specificity, and F1- score. The risk level associated with the URL will be assessed based on various factors, such as the source's reputation, content analysis results, and behavioural anomalies. The computational complexity of DL model training is influenced by various factors, such as the model's complexity, the training data's size, and the optimization algorithm exploited, for training. The outcome demonstrates that tweaking variables increases the effectiveness of Python-based deep learning systems. The findings of the proposed method excel, achieving an accuracy of 99.4 %, precision of 99.2 %, recall of 99.3 %, and an F1-score of 99.2 %. This innovative automatic phishing website detection and prevention model, based on a Transformer-based Deep Belief Network, offers advanced accuracy and adaptability, strengthening cybersecurity measures to safeguard sensitive user information and mitigate the substantial threat of phishing attacks in the digitally connected world.

Research on a soft saturation nonlinear SSVEP signal feature extraction algorithm

Brain–computer interfaces (BCIs) based on steady-state visual evoked potentials (SSVEP) have received widespread attention due to their high information transmission rate, high accuracy, and rich instruction set. However, the performance of its identification methods strongly depends on the amount of calibration data for within-subject classification. Some studies use deep learning (DL) algorithms for inter-subject classification, which can reduce the calculation process, but there is still much room for improvement in performance compared with intra-subject classification. To solve these problems, an efficient SSVEP signal recognition deep learning network model e-SSVEPNet based on the soft saturation nonlinear module is proposed in this paper. The soft saturation nonlinear module uses a similar exponential calculation method for output when it is less than zero, improving robustness to noise. Under the conditions of the SSVEP data set, two sliding time window lengths (1 s and 0.5 s), and three training data sizes, this paper evaluates the proposed network model and compares it with other traditional and deep learning model baseline methods. The experimental results of the nonlinear module were classified and compared. A large number of experimental results show that the proposed network has the highest average accuracy of intra-subject classification on the SSVEP data set, improves the performance of SSVEP signal classification and recognition, and has higher decoding accuracy under short signals, so it has huge potential ability to realize high-speed SSVEP-based for BCI.

Data-driven prediction of tool wear using Bayesian regularized artificial neural networks

The prediction of wear in cutting tools is pivotal for boosting productivity and reducing manufacturing costs. Although current data-driven models in machine learning and deep learning have advanced predictive capabilities, they often lack generality and demand substantial data training. This paper presents a novel approach using Bayesian Regularized Artificial Neural Networks (BRANNs) to precisely forecast wear in milling tools. Unlike conventional machine learning models, BRANNs merge the strengths of artificial neural networks (ANNs) and Bayesian regularization, yielding a more robust and generalized predictive model. We utilized three open-access datasets from the literature alongside an in-house dataset generated by our milling setup. Initially, we assessed the model’s predictive ability by training and testing it against individual open-access datasets. We investigated the impact of input features, training data size, hidden units, training algorithms, and transfer functions on the model’s predictive capability. Subsequently, we trained the model using three open-access datasets and tested it against our in-house data. Our findings demonstrate that the developed model surpasses existing state-of-the-art models in accuracy and transferability

Learning linear optical circuits with coherent states

We analyze the energy and training data requirements for supervised learning of an M-mode linear optical circuit by minimizing an empirical risk defined solely from the action of the circuit on coherent states. When the linear optical circuit acts non-trivially only on k < M unknown modes (i.e. a linear optical k-junta), we provide an energy-efficient, adaptive algorithm that identifies the junta set and learns the circuit. We compare two schemes for allocating a total energy, E, to the learning algorithm. In the first scheme, each of the T random training coherent states has energy E/T . In the second scheme, a single random MT-mode coherent state with energy E is partitioned into T training coherent states. The latter scheme exhibits a polynomial advantage in training data size sufficient for convergence of the empirical risk to the full risk due to concentration of measure on the (2MT−1) -sphere. Specifically, generalization bounds for both schemes are proven, which indicate that for ε-approximation of the full risk by the empirical risk with high probability, O(E2/3M2/3/ϵ2/3) training states are sufficient for the first scheme and O(E1/3M1/3/ϵ2/3) training states are sufficient for the second scheme.