Terms Of Training Time Research Articles

Transfer of training—Virtual reality training with augmented multisensory cues improves user experience during training and task performance in the real world

Virtual reality (VR) can create safe, cost-effective, and engaging learning environments. It is commonly assumed that improvements in simulation fidelity lead to better learning outcomes. Some aspects of real environments, for example vestibular or haptic cues, are difficult to recreate in VR, but VR offers a wealth of opportunities to provide additional sensory cues in arbitrary modalities that provide task relevant information. The aim of this study was to investigate whether these cues improve user experience and learning outcomes, and, specifically, whether learning using augmented sensory cues translates into performance improvements in real environments. Participants were randomly allocated into three matched groups: Group 1 (control) was asked to perform a real tyre change only. The remaining two groups were trained in VR before performance was evaluated on the same, real tyre change task. Group 2 was trained using a conventional VR system, while Group 3 was trained in VR with augmented, task relevant, multisensory cues. Objective performance, time to completion and error number, subjective ratings of presence, perceived workload, and discomfort were recorded. The results show that both VR training paradigms improved performance for the real task. Providing additional, task-relevant cues during VR training resulted in higher objective performance during the real task. We propose a novel method to quantify the relative performance gains between training paradigms that estimates the relative gain in terms of training time. Systematic differences in subjective ratings that show comparable workload ratings, higher presence ratings and lower discomfort ratings, mirroring objective performance measures, were also observed. These findings further support the use of augmented multisensory cues in VR environments as an efficient method to enhance performance, user experience and, critically, the transfer of training from virtual to real environment scenarios.

Relation extraction for coal mine safety information using recurrent neural networks with bidirectional minimal gated unit

The data of coal mine safety field are massive, multi-source and heterogeneous. It is of practical importance to extract information from big data to achieve disaster precaution and emergency response. Existing approaches need to build more features and rely heavily on the linguistic knowledge of researchers, leading to inefficiency, poor portability, and slow update speed. This paper proposes a new relation extraction approach using recurrent neural networks with bidirectional minimal gated unit (MGU) model. This is achieved by adding a back-to-front MGU layer based on original MGU model. It does not require to construct complex text features and can capture the global context information by combining the forward and backward features. Evident from extensive experiments, the proposed approach outperforms the existing initiatives in terms of training time, accuracy, recall and F value.

On Model Selection for Scalable Time Series Forecasting in Transport Networks

The transport literature is dense regarding short-term traffic predictions, up to the scale of 1 hour, yet less dense for long-term traffic predictions. The transport literature is also sparse when it comes to city-scale traffic predictions, mainly because of low data availability. In this work, we report an effort to investigate whether deep learning models can be useful for the long-term large-scale traffic prediction task, while focusing on the scalability of the models. We investigate a city-scale traffic dataset with 14 weeks of speed observations collected every 15 minutes over 1098 segments in the hypercenter of Los Angeles, California. We look at a variety of state-of-the-art machine learning and deep learning predictors for link-based predictions, and investigate how such predictors can scale up to larger areas with clustering, and graph convolutional approaches. We discuss that modelling temporal and spatial features into deep learning predictors can be helpful for long-term predictions, while simpler, not deep learning-based predictors, achieve very satisfactory performance for link-based and short-term forecasting. The trade-off is discussed not only in terms of prediction accuracy vs prediction horizon but also in terms of training time and model sizing.

Training with varying odor concentrations: implications for odor detection thresholds in canines.

Detection dogs are required to detect trace quantities of substances, many times in the parts per billion or parts per trillion concentration range. Frequently, detection of trace quantities is not explicitly trained but rather assumed when dogs show proficiency at higher concentrations to which they are trained. The aim of this study was to evaluate the effect of the odor concentration of the training sample on the minimum concentration dogs will subsequently detect. We expected that dogs may not spontaneously generalize to trace odor concentration when trained with higher concentrations, but when trained to a range of lower concentrations, dogs will show superior detection to lower untrained concentrations. A total of 11 dogs were randomly assigned to 2 groups and were trained to alert to isoamyl acetate at 0.01% odor dilution (v/v with mineral oil) using a 3-alternative forced choice test. Once reaching proficiency, odor detection threshold was assessed using a 2-down 1-up descending staircase procedure. Next, experimental dogs received training with systematically lower concentrations of isoamyl acetate and threshold re-assessed. Control dogs were yoked to experimental dogs in terms of training time, but only received training to the 0.01% dilution between threshold assessments. Experimental dogs showed significantly improved detection thresholds, outperforming control dogs by detecting an average dilution about 100-fold lower. Results suggest that explicitly training for lower concentrations is critical for generalization for trace odor detection.

Effects of Training Set Size on Supervised Machine-Learning Land-Cover Classification of Large-Area High-Resolution Remotely Sensed Data

The size of the training data set is a major determinant of classification accuracy. Nevertheless, the collection of a large training data set for supervised classifiers can be a challenge, especially for studies covering a large area, which may be typical of many real-world applied projects. This work investigates how variations in training set size, ranging from a large sample size (n = 10,000) to a very small sample size (n = 40), affect the performance of six supervised machine-learning algorithms applied to classify large-area high-spatial-resolution (HR) (1–5 m) remotely sensed data within the context of a geographic object-based image analysis (GEOBIA) approach. GEOBIA, in which adjacent similar pixels are grouped into image-objects that form the unit of the classification, offers the potential benefit of allowing multiple additional variables, such as measures of object geometry and texture, thus increasing the dimensionality of the classification input data. The six supervised machine-learning algorithms are support vector machines (SVM), random forests (RF), k-nearest neighbors (k-NN), single-layer perceptron neural networks (NEU), learning vector quantization (LVQ), and gradient-boosted trees (GBM). RF, the algorithm with the highest overall accuracy, was notable for its negligible decrease in overall accuracy, 1.0%, when training sample size decreased from 10,000 to 315 samples. GBM provided similar overall accuracy to RF; however, the algorithm was very expensive in terms of training time and computational resources, especially with large training sets. In contrast to RF and GBM, NEU, and SVM were particularly sensitive to decreasing sample size, with NEU classifications generally producing overall accuracies that were on average slightly higher than SVM classifications for larger sample sizes, but lower than SVM for the smallest sample sizes. NEU however required a longer processing time. The k-NN classifier saw less of a drop in overall accuracy than NEU and SVM as training set size decreased; however, the overall accuracies of k-NN were typically less than RF, NEU, and SVM classifiers. LVQ generally had the lowest overall accuracy of all six methods, but was relatively insensitive to sample size, down to the smallest sample sizes. Overall, due to its relatively high accuracy with small training sample sets, and minimal variations in overall accuracy between very large and small sample sets, as well as relatively short processing time, RF was a good classifier for large-area land-cover classifications of HR remotely sensed data, especially when training data are scarce. However, as performance of different supervised classifiers varies in response to training set size, investigating multiple classification algorithms is recommended to achieve optimal accuracy for a project.

Image reconstruction through a multimode fiber with a simple neural network architecture

Multimode fibers (MMFs) have the potential to carry complex images for endoscopy and related applications, but decoding the complex speckle patterns produced by mode-mixing and modal dispersion in MMFs is a serious challenge. Several groups have recently shown that convolutional neural networks (CNNs) can be trained to perform high-fidelity MMF image reconstruction. We find that a considerably simpler neural network architecture, the single hidden layer dense neural network, performs at least as well as previously-used CNNs in terms of image reconstruction fidelity, and is superior in terms of training time and computing resources required. The trained networks can accurately reconstruct MMF images collected over a week after the cessation of the training set, with the dense network performing as well as the CNN over the entire period.

Text classification by untrained sentence embeddings

Recurrent Neural Networks (RNNs) represent a natural paradigm for modeling sequential data like text written in natural language. In fact, RNNs and their variations have long been the architecture of choice in many applications, however in practice they require the use of labored architectures (such as gating mechanisms) and computationally heavy training processes. In this paper we address the question of whether it is possible to generate sentence embeddings via completely untrained recurrent dynamics, on top of which to apply a simple learning algorithm for text classification. This would allow to obtain extremely efficient models in terms of training time. Our work investigates the extent to which this approach can be used, by analyzing the results on different tasks. Finally, we show that, within certain limits, it is possible to build extremely efficient models for text classification that remain competitive in accuracy with reference models in the state-of-the-art.

Feature-Rich Classifiers for Recognizing Textual Entailment in Indonesian

Recognizing Textual Entailment (RTE) is a Natural Language Processing task to determine whether a sentence (text) semantically entails another sentence (hypothesis). In this paper, we extracted and learned 35 features from a pair of text and hypothesis in Indonesian. The ablation study was conducted to analyze features contribution to RTE model. The experiments shown that, using Support Vector Machine (SVM) and Logistic Regression, the token-based features contribute positively to improve the model performance. The best model in our experiment is SVM that scored F1-Score of 79.65%. Despite sacrificing 5-points accuracy to the state-of-the-art BERT model, SVM classifier is 31 hours more efficient in terms of training time.

Multi-Disease Classification Model Using Strassen’s Half of Threshold (SHoT) Training Algorithm in Healthcare Sector

In healthcare industry, Neural Network has attained a milestone in solving many real-life classification problems varies from very simple to complex and from linear to non-linear. To improve the training process by reducing the training time, Adaptive Skipping Training algorithm named as Half of Threshold (HOT) has been proposed. To perform the fast classification and also to improve the computational efficiency such as accuracy, error rate, etc., the highlighted characteristics of proposed HOT algorithm has been integrated with Strassen’s matrix multiplication algorithm and derived a novel, hybrid and computationally efficient algorithm for training and validating the neural network named as Strassen’s Half of Threshold (SHoT) Training Algorithm. The experimental outcome based on the simulation demonstrated that the proposed SHOT algorithm outperforms both BPN and HOT algorithm in terms of training time which is reduced with the range of 7% to 54% and its efficiency which is improved with the range of 3% to 15% on various dataset such as Hepatitis, SPeCT, Heart, Liver Disorders, Breast Cancer Wisconsin (Diagnostic), Drug Consumption, Cardiotocography, Splice-junction Gene Sequences and Thyroid Disease dataset that are extracted from Machine Learning Dataset Repository of UCI. It can be integrated with any type of supervised training algorithm.

Comparison analysis of decision tree and ensemble models in the classification of chronic kidney diseases

The world is now in the era of big data and processing, and exploring the data has become one of the significant challenges. Hence, researchers have done a lot to analyse these data in the health sector to enhance disease detection and classification using artificial intelligence and ML principles. Kidney disease is one of the terrible conditions in which its late detection has sent many people to untimely graves. ML classifiers have been employed in many dimensions to classify heart disease, but, existing works have not explored the variants of each method for selection of best model parameters. An attempt is being made in this research to study the behaviour of three (3) variants each from two(2) tree-based models in the classification of Kidney Disease. Three of the variants are Complex, Medium and Simple models of Decision tree classifier and the other one are Boosted, Bagged and RUSBoosted of Ensemble Classifiers. Using MATLAB for implementation, the model performance established that the accuracy of Ensemble Classifier (Bagged tree model) is the best, concerning the speed, Decision tree (Complex and Simple tree models have the same and highest value). Hence, the two are the best. In terms of training time, Decision tree(Simple tree) has the least time and therefore the best.

Medical image segmentation and reconstruction of prostate tumor based on 3D AlexNet

BackgroundProstate cancer is a disease with a high incidence of tumors in men. Due to the long incubation time and insidious condition, early diagnosis is difficult; especially imaging diagnosis is more difficult. In actual clinical practice, the method of manual segmentation by medical experts is mainly used, which is time-consuming and labor-intensive and relies heavily on the experience and ability of medical experts. The rapid, accurate and repeatable segmentation of the prostate area is still a challenging problem. It is important to explore the automated segmentation of prostate images based on the 3D AlexNet network. MethodTaking the medical image of prostate cancer as the entry point, the three-dimensional data is introduced into the deep learning convolutional neural network. This paper proposes a 3D AlexNet method for the automatic segmentation of prostate cancer magnetic resonance images, and the general network ResNet 50, Inception -V4 compares network performance. ResultsBased on the training samples of magnetic resonance images of 500 prostate cancer patients, a set of 3D AlexNet with simple structure and excellent performance was established through adaptive improvement on the basis of classic AlexNet. The accuracy rate was as high as 0.921, the specificity was 0.896, and the sensitivity It is 0.902 and the area under the receiver operating characteristic curve (AUC) is 0.964. The Mean Absolute Distance (MAD) between the segmentation result and the medical expert's gold standard is 0.356 mm, and the Hausdorff distance (HD) is 1.024 mm, the Dice similarity coefficient is 0.9768. ConclusionThe improved 3D AlexNet can automatically complete the structured segmentation of prostate magnetic resonance images. Compared with traditional segmentation methods and depth segmentation methods, the performance of the 3D AlexNet network is superior in terms of training time and parameter amount, or network performance evaluation. Compared with the algorithm, it proves the effectiveness of this method.

Multi-view Incremental Discriminant Analysis

In recent years, the use of multi-view data has attracted much attention resulting in many multi-view batch learning algorithms. However, these algorithms prove expensive in terms of training time and memory when used on the incremental data. In this paper, we propose Multi-view Incremental Discriminant Analysis (MvIDA), which updates the trained model to incorporate new data samples. MvIDA requires only the old model and newly added data to update the model. Depending on the nature of the increments, MvIDA is presented as two cases, sequential MvIDA and chunk MvIDA. We have compared the proposed method against the batch Multi-view Discriminant Analysis (MvDA) for its discriminability, order independence, the effect of the number of views, training time, and memory requirements. We have also compared our method with single-view Incremental Linear Discriminant Analysis (ILDA) for accuracy and training time. The experiments are conducted on four datasets with a wide range of dimensions per view. The results show that through order independence and faster construction of the optimal discriminant subspace, MvIDA addresses the issues faced by the batch multi-view algorithms in the incremental setting.

Low-rank tensor ring learning for multi-linear regression

The emergence of large-scale data demands new regression models with multi-dimensional coefficient arrays, known as tensor regression models. The recently proposed tensor ring decomposition has interesting properties of enhanced representation and compression capability, cyclic permutation invariance and balanced tensor ring rank, which may lead to efficient computation and fewer parameters in regression problems. In this paper, a generally multi-linear tensor-on-tensor regression model is proposed that the coefficient array has a low-rank tensor ring structure, which is termed tensor ring ridge regression (TRRR). Two optimization models are developed for the TRRR problem and solved by different algorithms: the tensor factorization based one is solved by alternating least squares algorithm, and accelerated by a fast network contraction, while the rank minimization based one is addressed by the alternating direction method of multipliers algorithm. Comparative experiments, including Spatio-temporal forecasting tasks and 3D reconstruction of human motion capture data from its temporally synchronized video sequences, demonstrate the enhanced performance of our algorithms over existing state-of-the-art ones, especially in terms of training time.

Fast Channel Estimation for IRS-Assisted OFDM

In this letter, we study efficient channel estimation for an intelligent reflecting surface (IRS)-assisted orthogonal frequency division multiplexing (OFDM) system to achieve minimum training time. First, a fast channel estimation scheme with reduced OFDM symbol duration is proposed for arbitrary frequency-selective fading channels. Next, under the typical condition that the IRS-user channel is line-of-sight (LoS) dominant, another fast channel estimation scheme based on the novel concept of sampling-wise IRS reflection variation is proposed. Moreover, the pilot signal and IRS training reflection pattern are jointly optimized for both proposed schemes. Finally, the proposed schemes are compared in terms of training time and channel estimation performance via simulations, as well as against benchmark schemes.

Prediction of Fatigue Crack Growth Behaviour in Ultrafine Grained Al 2014 Alloy Using Machine Learning

The present work investigates the relationship between fatigue crack growth rate (da/dN) and stress intensity factor range (∆K) using machine learning models with the experimental fatigue crack growth rate (FCGR) data of cryo-rolled Al 2014 alloy. Various machine learning techniques developed recently provide a flexible and adaptable approach to explain the complex mathematical relations especially, non-linear functions. In the present work, three machine algorithms such as extreme learning machine (ELM), back propagation neural networks (BPNN) and curve fitting model are implemented to analyse FCGR of Al alloys. After tuning of networks with varying hidden layers and number of neurons, the trained models found to fit well to the tested data. The three tested models are compared with each other over the training as well as testing phase. The mean square error for predicting the FCG of cryo-rolled Al 2014 alloy by BPNN, ELM and curve fitting methods are 1.89, 1.84 and 0.09 respectively. While the ELM models outperform the rest of models in terms of training time, curve fitting model showed best performance in terms of accuracy over testing data with least mean square error (MSE). In terms of local optimisation, back propagation neural networks excel the other two models.

Deep kernel learning in extreme learning machines

Emergence of extreme learning machine as a breakneck learning algorithm has marked its prominence in solitary hidden layer feed-forward networks. Kernel-based extreme learning machine (KELM) reflected its efficiency in diverse applications where feature mapping functions of hidden nodes are concealed from users. The conventional KELM algorithms involve only solitary layer of kernels, thereby emulating shallow learning architectures for its feature transformation. Trend in migrating shallow-based learning models into deep learning architectures opens up a new outlook for machine learning domains. This paper attempts to bestow deep kernel learning approach in a conventional shallow architecture. The emerging arc-cosine kernels possess the potential to mimic the prevailing deep layered frameworks to a greater extent. Unlike other kernels such as linear, polynomial and Gaussian, arc-cosine kernels have a recursive nature by itself and have the potential to express multilayer computation in learning models. This paper explores the possibility of building a new deep kernel machine with extreme learning machine and multilayer arc-cosine kernels. This framework outperforms conventional KELM and deep support vector machine in terms of training time and accuracy.

Vibrotactile Alphabets: Time and Frequency Patterns to Encode Information.

Although many tactual communication methods have been developed over the past decades, few of them have been widely adopted in everyday lives. In this paper, we investigate the ability to deliver text messages using patterns of vibrations. By manipulating the duration and frequency of vibrations, we designed two vibrotactile alphabets, i.e. 3F Vibraille and 2F Vibraille, along with their corresponding numbers and symbols. The main difference between 3F Vibraille and 2F Vibraille is the number of frequency levels utilised. Both alphabets were evaluated through human experiments in terms of training time, learning process and recognition accuracy of words and symbols. For both alphabets, the results demonstrated that participants were able to recognise words and symbols at an accuracy of over 90% after 6-8 hours of training. The main difference between the alphabets was that 3F Vibraille achieved a higher transmission rate while 2F Vibraille required less training time.

Analyzing MRI scans to detect glioblastoma tumor using hybrid deep belief networks

Glioblastoma (GBM) is a stage 4 malignant tumor in which a large portion of tumor cells are reproducing and dividing at any moment. These tumors are life threatening and may result in partial or complete mental and physical disability. In this study, we have proposed a classification model using hybrid deep belief networks (DBN) to classify magnetic resonance imaging (MRI) for GBM tumor. DBN is composed of stacked restricted Boltzmann machines (RBM). DBN often requires a large number of hidden layers that consists of large number of neurons to learn the best features from the raw image data. Hence, computational and space complexity is high and requires a lot of training time. The proposed approach combines DTW with DBN to improve the efficiency of existing DBN model. The results are validated using several statistical parameters. Statistical validation verifies that the combination of DTW and DBN outperformed the other classifiers in terms of training time, space complexity and classification accuracy.

An online isotonic separation with cascade architecture for binary classification

Isotonic separation (IS) is a non-parametric classification technique which constructs an isotonic function from ordered data. The rationale is to convert partially isotonic data into isotonic using a linear programming problem (LPP) and partition the input space into isotonic and non-isotonic regions to make predictions easier. Despite the widespread applications of IS in diverse domains where monotonicity exists, it has certain limitations: Firstly, computing time and the constraints of the LPP in isotonic separation increase polynomially as size of the data increases and it is highly complex to solve the LPP and obtain the model on large data sets. In order to support dynamic stream data and address the computational overhead and size of the LPP issues, this paper proposes an online isotonic separation algorithm called Cascade-IS (CIS) for binary classification. The rationale behind CIS is that it splits the data set into a sequence of partitions and models are obtained and combined in cascade. Statistical and experimental analysis are done on datasets with isotonic properties and the results prove that CIS is superior to its variants in terms of training time, performance measures and number of constraints in the LPP.

Machine Learning with ROOT/TMVA

ROOT provides, through TMVA, machine learning tools for data analysis at HEP experiments and beyond. We present recently included features in TMVA and the strategy for future developments in the diversified machine learning landscape. Focus is put on fast machine learning inference, which enables analysts to deploy their machine learning models rapidly on large scale datasets. The new developments are paired with newly designed C++ and Python interfaces supporting modern C++ paradigms and full interoperability in the Python ecosystem. We present as well a new deep learning implementation for convolutional neural network using the cuDNN library for GPU. We show benchmarking results in term of training time and inference time, when comparing with other machine learning libraries such as Keras/Tensorflow.