Redundant Instances Research Articles

Industry Foundation Class-Based Building Information Modeling Lightweight Visualization Method for Steel Structures

The efficient extraction, storage, and visualization of geometric and semantic information is a key foundation for the operation of the building information modeling (BIM) platform. This study aims to develop a lightweight BIM system and optimize the system’s performance according to the specific characteristics of steel structures. This study proposes several novel techniques for extracting and decoupling the geometric and semantic information of components from industry foundation class (IFC) files. A redundancy removal approach combining the principal content analysis (PCA) algorithm and the Hausdorff-based comparison algorithm is proposed to identify standardized steel components, and a lightweight visualization method on Web3D for redundant instances is also presented. A loading mechanism of the level of detail (LOD) model based on a mesh simplification algorithm is presented to optimize the display efficiency. The developed system is evaluated by three steel structural models. Using the redundancy removal approach, the number of instances is decreased by 96.46% in less than 30 s and over 30 FPS (frame per second) is kept when rendering. Using the LOD loading mechanism, 95.38% of vertices and 98.46% of patches are eliminated under 50 mm precision. The experiment results indicate that users can quickly load large BIM models and fetch sufficient information from the website.

An Item Response Theory Analysis of the Clinician-Administered PTSD Scale for DSM-5 Among Veterans.

We used item response theory (IRT) analysis to examine Clinician-Administered PTSD Scale for DSM-5 (CAPS-5) item performance using data from three large samples of veterans (total N = 808) using both binary and ordinal rating methods. Relative to binary ratings, ordinal ratings provided good coverage from well below to well above average within each symptom cluster. However, coverage varied by cluster, and item difficulties were unevenly distributed within each cluster, with numerous instances of redundancy. For both binary and ordinal scores, flashbacks, dissociative amnesia, and self-destructive behavior items showed a pattern of high difficulty but relatively poor discrimination. Results indicate that CAPS-5 ordinal ratings provide good severity coverage and that most items accurately differentiated between participants by severity. Observed uneven distribution and redundancy in item difficulty suggest there is opportunity to create an abbreviated version of the CAPS-5 for determining PTSD symptom severity, but not DSM-5 PTSD diagnosis, without sacrificing precision.

Notch pathway mutants do not equivalently perturb mouse embryonic retinal development.

In the vertebrate eye, Notch ligands, receptors, and ternary complex components determine the destiny of retinal progenitor cells in part by regulating Hes effector gene activity. There are multiple paralogues for nearly every node in this pathway, which results in numerous instances of redundancy and compensation during development. To dissect such complexity at the earliest stages of eye development, we used seven germline or conditional mutant mice and two spatiotemporally distinct Cre drivers. We perturbed the Notch ternary complex and multiple Hes genes to understand if Notch regulates optic stalk/nerve head development; and to test intracellular pathway components for their Notch-dependent versus -independent roles during retinal ganglion cell and cone photoreceptor competence and fate acquisition. We confirmed that disrupting Notch signaling universally blocks progenitor cell growth, but delineated specific pathway components that can act independently, such as sustained Hes1 expression in the optic stalk/nerve head. In retinal progenitor cells, we found that among the genes tested, they do not uniformly suppress retinal ganglion cell or cone differentiation; which is not due differences in developmental timing. We discovered that shifts in the earliest cell fates correlate with expression changes for the early photoreceptor factor Otx2, but not with Atoh7, a factor required for retinal ganglion cell formation. During photoreceptor genesis we also better defined multiple and simultaneous activities for Rbpj and Hes1 and identify redundant activities that occur downstream of Notch. Given its unique roles at the retina-optic stalk boundary and cone photoreceptor genesis, our data suggest Hes1 as a hub where Notch-dependent and -independent inputs converge.

A Comparative Survey of Instance Selection Methods applied to Non-Neural and Transformer-Based Text Classification

Progress in natural language processing has been dictated by the rule of more : more data, more computing power, more complexity, best exemplified by deep learning Transformers. However, training (or fine-tuning) large dense models for specific applications usually requires significant amounts of computing resources. One way to ameliorate this problem is through data engineering instead of the algorithmic or hardware perspectives. Our focus here is an under-investigated data engineering technique, with enormous potential in the current scenario – Instance Selection (IS) (a.k.a. Selective Sampling, Prototype Selection). The IS goal is to reduce the training set size by removing noisy or redundant instances while maintaining or improving the effectiveness (accuracy) of the trained models and reducing the training process cost. We survey classical and recent state-of-the-art IS techniques and provide a scientifically sound comparison of IS methods applied to an essential natural language processing task—Automatic Text Classification (ATC). IS methods have been normally applied to small tabular datasets and have not been systematically compared in ATC. We consider several neural and non-neural state-of-the-art ATC solutions and many datasets. We answer several research questions based on tradeoffs induced by a tripod (training set reduction, effectiveness, and efficiency). Our answers reveal an enormous unfulfilled potential for IS solutions. Specially, we show that in 12 out of 19 datasets, specific IS methods—namely, Condensed Nearest Neighbor, Local Set-based Smoother, and Local Set Border Selector—can reduce the size of the training set without effectiveness losses. Furthermore, in the case of fine-tuning the Transformer methods, the IS methods reduce the amount of data needed, without losing effectiveness and with considerable training-time gains.

On semantic redundancy in the epigraphs of Moesia Superior

AbstractIn this paper, the principle of semantic redundancy is examined as evidenced in the corpus of the Moesia Superior inscriptions (Inscriptions de la Mésie supérieure – IMS). The objective is to identify pleonastic phrases within the epigraphic material and to distinguish instances of unmotivated redundancy. The starting point is the observation made by Hofmann, Löfstedt, Herman, and others that this type of redundancy, contrary to the pragmatically motivated or “rhetoric” one, could reflect the ongoing semantic change in Vulgar Latin. Given the scarcity of semantic redundancy findings in the IMS corpus, this paper also serves as an introduction to this linguistically important topic.

Improvements on the reverse order laws

AbstractIn many papers concerning properties of generalized inverses in different settings, we can find the results with many redundant instances of assuming regularity of certain elements. We have made an effort to widen the range of applicability of concrete results by considering more general cases of the problems without imposing any such additional assumptions. This was the main motivation for this paper, and we present several significant improvements of the reverse order laws on and ‐inverses in the ring setting.

YaConv: Convolution with Low Cache Footprint

This article introduces YaConv , a new algorithm to compute convolution using GEMM microkernels from a Basic Linear Algebra Subprograms library that is efficient for multiple CPU architectures. Previous approaches either create a copy of each image element for each filter element or reload these elements into cache for each GEMM call, leading to redundant instances of the image elements in cache. Instead, YaConv loads each image element once into the cache and maximizes the reuse of these elements. The output image is computed by scattering results of the GEMM microkernel calls to the correct locations in the output image. The main advantage of this new algorithm—which leads to better performance in comparison to the existing im2col approach on several architectures—is a more efficient use of the memory hierarchy. The experimental evaluation on convolutional layers from PyTorch, along with a parameterized study, indicates an average 24% speedup over im2col convolution. Increased performance comes as a result of 3× reduction in L3 cache accesses and 2× fewer branch instructions.

The use of discourse markers in argumentative compositions by Jordanian EFL learners

The aim of the present study is to investigate the use of discourse markers (DMs) in the argumentative compositions written by EFL learners at two academic stages (sophomores and seniors) majoring in English at the Hashemite University, Jordan. The significance of this study springs from its focus on the use of DMs in Jordanian EFL learners’ argumentative writings. Employing an integrated research method of qualitative and quantitative analysis, the findings revealed that both groups of participants used the same types of DMs with varying degree of frequency, namely, elaborative, contrastive, reason, inferential, conclusive, and exemplifier DMs, respectively. The sophomores were observed to employ a relatively higher number of DMs compared to the seniors, which may be ascribed to some redundant instances of DMs. The elaborative, contrastive, and reason types were the most widely used, while inferentials, conclusives and exemplifiers appeared infrequently in both groups. The analysis of individual DMs displayed that the DMs ‘and’, ‘because’, and ‘but’ were the predominant across the seniors and sophomores’ argumentative texts. This overuse of these DMs may be due to the influence of L1 of the participants and the popularity of these DMs among students and teachers of English. Additionally, the participants showed a low proficiency in using DMs since they overused largely a restricted variety of DMs at the expense of others that would be expected in the argumentative writing; some DMs were noticed either to be underused or absent. The results of Pearson’s r correlation test indicated that there was a weak positive but significant correlation between the writing quality and the use of DMs. This may be taken as a predictor of the writing quality in argumentative compositions by EFL. Pedagogically, the study emphasizes the significance of teaching DMs, where EFL learners should be taught how to use them appropriately to avoid any transference of their L1. Further research on DMs in argumentative writings in different levels of proficiency is recommended.

Density-guided Incremental Dominant Instance Exploration for Two-view Geometric Model Fitting.

Existing two-view multi-model fitting methods typically follow a two-step manner, i.e., model generation and selection, without considering their interaction. Therefore, in the first step, these methods have to generate a considerable number of instances in order to cover all desired ones, which not only offers no guarantees, but also introduces unnecessary expensive calculations. To address this challenge, this study presents a new algorithm, termed as D2Fitting, that incrementally explores dominant instances. Particularly, rather than viewing model generation and selection as two disjoint parts, D2Fitting fully considers their interaction, and thus performs these two subroutines alternatively under a simple yet effective optimization framework. This design can avoid generating too many redundant instances, thus reducing computational overhead and allowing the proposed D2Fitting being real-time. Meanwhile, we further design a novel density-guided sampler to sample high-quality minimal subsets during the model generation process, so as to fully exploit the spatial distribution of the input data. Also, to mitigate the influence of noise on the subsets sampled by the proposed sampler, a global-residual optimization strategy is investigated for the minimal subset refinement. With all the ingredients mentioned above, the proposed D2Fitting can accurately estimate the number and parameters of geometric models and efficiently segment the input data simultaneously. Extensive experiments on several public datasets demonstrate the significant superiority of D2Fitting over several state-of-the-arts.

A hybrid data‐level sampling approach in learning from skewed user‐click data for click fraud detection in online advertising

AbstractOne of the challenging issues in user‐click data of online advertising is the uneven class distribution which biases classification models. Resampling the data is a popular choice for obtaining class balance. However, oversampling results in overfitting, whilst under‐sampling results in information loss. Moreover, enhancing separability between samples, where the classes overlap closer to the decision boundary, is another challenge, which requires a careful pruning of instances towards increasing the separability in data space. Therefore, in this work, a new hybrid data sampling algorithm SMOTEOSS is designed and evaluated, concatenating the synthetic minority oversampling technique (SMOTE) followed by one‐sided selection (OSS) to balance the class distribution. The working of SMOTEOSS is twofold‐ first, it oversamples the under‐represented class distribution using the SMOTE by generating synthetic instances. However, the generation of synthetic instances closer to the decision boundary directly influences the learning model's decision‐making. Utilising OSS, the proposed method then identifies TOMEKLINKS and eliminates the noisy majority instances whilst eliminating the redundant instances. The proposed method's effectiveness is validated on the FDMA 2012 dataset against 10 state‐of‐the‐art sampling methods utilising the gradient tree boosting learning model. To authenticate SMOTEOSS, a fair comparison is made by conducting experiments on other 10 benchmark imbalanced datasets using 10‐fold cross‐validation. Performance is measured using average precision, recall, F1‐score, G‐mean, the area under curve (AUC) and reduction rate. Results showed that the designed hybrid methodology is an efficient alternative to existing sampling methods. The Wilcoxon signed‐rank test is employed to demonstrate significant differences amidst the proposed and conventional sampling algorithms.

CoShare: An Efficient Approach for Redundancy Allocation in NFV

An appealing feature of Network Function Virtualization (NFV) is that in an NFV-based network, a network function (NF) instance may be placed at any node. On the one hand this offers great flexibility in allocation of redundant instances, but on the other hand it makes the allocation a unique and difficult challenge. One particular concern is that there is inherent correlation among nodes due to the structure of the network, thus requiring special care in this allocation. To this aim, our novel approach, called <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">CoShare</i> , is proposed. Firstly, its design takes into consideration the effect of network structural dependency, which might result in the unavailability of nodes of a network after failure of a node. Secondly, to efficiently make use of resources, CoShare proposes the idea of <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">shared reservation</i> , where multiple flows may be allowed to share the same reserved backup capacity at an NF instance. Furthermore, CoShare factors in the heterogeneity in nodes, NF instances and availability requirements of flows in the design. The results from a number of experiments conducted using realistic network topologies show that the integration of structural dependency allows meeting availability requirements for more flows compared to a baseline approach. Specifically, CoShare is able to meet diverse availability requirements in a resource-efficient manner, requiring, e.g., up to 85% in some studied cases, less resource overbuild than the baseline approach that uses the idea of <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">dedicated reservation</i> commonly adopted for redundancy allocation in NFV.

Translation and the Second Language Learner: Examining English Language Usage among Students of Federal University of Technology Owerri, Nigeria

This paper examined, using students of Federal University of Technology Owerri, Nigeria as case study, translation as a significant phenomenon in the Igbo language native speaker’s use of English as a second language. The purpose was to identify the manners of such translation, and the possible motivations. A total of two hundred and four (204) students admitted into the School of Health Technology of the University for the 2020/2021 academic year formed the sample population. The method of data collection was purposive as the authors paid particular attention to instances of direct translation as they manifested in the students’ discussions and interactions. The students were not subjected to any form of controlled environment, but rather, the authors as lecturers of the selected students engaged in direct observation of the students in their social interactions as well as formal class discussions. The items of expressions extracted as data were subjected to evaluation with the aim of identifying incidences and manner of direct translation from Igbo language structure to the English language structure, as well as their possible sources of motivation. Thus, the method of data analysis was qualitative. The paper adopted the theoretical approach of word-for-word translation or metaphrasing. The findings of the paper include that the affected students think in the source language (SL) before translating their thoughts into the target language (TL). What is experienced as TL structures follows from the SL structures that exist in the mental resources of the students. The paper, therefore, concludes that direct translation can be deceptive to the second language learner because it may not contain obvious instances of structural rule deficiency but can be faulted along functional lane, where it manifests instances of redundancy, and vocabulary deficiency. Consequently, the authors recommend among others that the L2 teacher should extend his/ her teaching efforts towards identifying the possible motivation a student’s direct translation in order to guide him or her out of it, especially, the students who may need the language beyond its L2 environment, or beyond mere language of social expression.

Features Selection for Supervised Learning Using Centrality Measures

The data mining methods have been extensively used in the process of decision making. The popularity of data mining methods is due to availability of high speed algorithms, processing and storage power of computers. The effective use of data mining methods help in mining datasets and taking better decisions. The data need to be preprocessed before applying data mining methods. Some datasets require little preparation like dealing with missing and redundant instances while some high-dimensional datasets require strong processing like dimensionality reduction. One of the techniques used for dimensionality reduction is feature selection. This study uses graph based centrality measure for feature selection. Graph based centrality measures are used for ranking features which is used for removing irrelevant attributes. After comparison of results with other approaches, it has been found that the proposed approach results in reduction of feature space without compromising accuracy. The results also shows that proposed approach performs better than some other feature selection approaches not only in terms of accuracy but also on the basis of larger reduction in feature space.

Dual Data Selection Using Multi-Objective Micro-CHC

Objective: Redundant and superfluous features or instances reduce the efficiency and efficacy of data mining algorithms. Hence, selecting relevant and significant features and instances is very important for a data mining process to be able to discern some meaning information. Dual selection deals with the problem of generating a small subset of non-redundant features as well as instances simultaneously from a large and noisy data set. The two main objectives for dual selection are to maximize the classification accuracy and to bring as much as possible data reduction. The two objectives, accuracy and data reduction rate, are conflicting because maximizing the data reduction rate generally results in a lower accuracy rate and vice versa. These objectives are mutually dependent and must be tackled simultaneously. Therefore, the problem of dual data selection ought to be naturally approached with multi-objective optimization techniques which give a set of nondominated solutions instead of a single best solution. The problem of dual selection has exhaustively large search space and has been addressed through single and Multi-Objective Genetic Algorithms (MOGAs). More often, evolutionary approaches may it be single or multi-objective work with large population sizes and take unacceptably long execution times due to computationally expensive fitness functions. These approaches also suffer from premature convergence. Methods: This paper proposes a hybrid Multi-Objective Micro-CHC (MO-Micro-CHC) to address the task of dual selection. The suggested approach uses a population of only a few individuals and elitism advised in Micro Genetic Algorithm (Micro-GA), Heterogeneous Uniform Recombination (HUX) and Cataclysmic mutation inspired by CHC, and non-dominated sorting of NSGA-II- a most popular and widely implemented multi-objective genetic algorithm. Results: We have conducted extensive experimentation using numerous datasets from the UCI data repository. Analysis of results approves that Mo-Micro-CHC achieves high accuracy and competing reduction rate in comparison to similar approaches. In addition, it takes far less execution time as compared to many of its counterparts.

Firefly algorithm for instance selection

The paper focuses on the problem of instance selection. Instance selection is currently crucial to enhance the efficacy and efficiency of machine-learning tools when they are used to solve a data-mining task and when the data are large and they are seen through the prism of the big data phenomenon. Instance selection eliminates redundant instances and thus reduces the size of the training data set. The training data, with redundant cases removed, can be more useful and ensure better performance of the final classification models. The instance selection problem belongs to the NP-hard class, so it can be solved with an approximation tool. In this paper the firefly algorithm is proposed for solving the instance selection problem. This paper is one paper, where the firefly algorithm has been used to solve a discrete optimisation problem, when in more cases previously it has been used for solving continuous optimisation problems. The firefly-based instance selection algorithm is presented and its validation is carried out. The results of the computational experiment show that the algorithm is competitive with others. The results obtained are discussed and conclusions are formulated.

Spontaneous homeotic mutants and genetic control of floral organ identity in a ranunculid.

The regulation of floral organ identity was investigated using a forward genetic approach in five floral homeotic mutants of Thalictrum, a noncore eudicot. We hypothesized that these mutants carry defects in the floral patterning genes. Mutant characterization comprised comparative floral morphology and organ identity gene expression at early and late developmental stages, followed by sequence analysis of coding and intronic regions to identify transcription factor binding sites and protein-protein interaction (PPI) motifs. Mutants exhibited altered expression of floral MADS-box genes, which further informed the function of paralogs arising from gene duplications not found in reference model systems. The ensuing modified BCEmodels for the mutants supported instances of neofunctionalization (e.g., B-class genes expressed ectopically in sepals), partial redundancy (E-class), or subfunctionalization (C-class) of paralogs. A lack of deleterious mutations in the coding regions of candidate floral MADS-box genes suggested that cis-regulatory or trans-acting mutations are at play. Consistent with this hypothesis, double-flower mutants had transposon insertions or showed signs of transposon activity in the regulatory intron of AGAMOUS (AG) orthologs. Single amino acid substitutions were also found, yet they did not fall on any of the identified DNA binding or PPI motifs.In conclusion, we present evidence suggesting that transposon activity and regulatory mutations in floral homeotic genes likely underlie the striking phenotypes of these Thalictrum floral homeotic mutants.

Hybrid Harris hawks optimization with cuckoo search for drug design and discovery in chemoinformatics

One of the major drawbacks of cheminformatics is a large amount of information present in the datasets. In the majority of cases, this information contains redundant instances that affect the analysis of similarity measurements with respect to drug design and discovery. Therefore, using classical methods such as the protein bank database and quantum mechanical calculations are insufficient owing to the dimensionality of search spaces. In this paper, we introduce a hybrid metaheuristic algorithm called CHHO–CS, which combines Harris hawks optimizer (HHO) with two operators: cuckoo search (CS) and chaotic maps. The role of CS is to control the main position vectors of the HHO algorithm to maintain the balance between exploitation and exploration phases, while the chaotic maps are used to update the control energy parameters to avoid falling into local optimum and premature convergence. Feature selection (FS) is a tool that permits to reduce the dimensionality of the dataset by removing redundant and non desired information, then FS is very helpful in cheminformatics. FS methods employ a classifier that permits to identify the best subset of features. The support vector machines (SVMs) are then used by the proposed CHHO–CS as an objective function for the classification process in FS. The CHHO–CS-SVM is tested in the selection of appropriate chemical descriptors and compound activities. Various datasets are used to validate the efficiency of the proposed CHHO–CS-SVM approach including ten from the UCI machine learning repository. Additionally, two chemical datasets (i.e., quantitative structure-activity relation biodegradation and monoamine oxidase) were utilized for selecting the most significant chemical descriptors and chemical compounds activities. The extensive experimental and statistical analyses exhibit that the suggested CHHO–CS method accomplished much-preferred trade-off solutions over the competitor algorithms including the HHO, CS, particle swarm optimization, moth-flame optimization, grey wolf optimizer, Salp swarm algorithm, and sine–cosine algorithm surfaced in the literature. The experimental results proved that the complexity associated with cheminformatics can be handled using chaotic maps and hybridizing the meta-heuristic methods.

Particle Swarm Optimization Based Swarm Intelligence for Active Learning Improvement: Application on Medical Data Classification

Semi-supervised learning targets the common situation where labeled data are scarce but unlabeled data are abundant. It uses unlabeled data to help supervised learning tasks. In practice, it may make sense to utilize active learning in conjunction with semi-supervised learning. That is, we might allow the learning algorithm to pick a set of unlabeled instances to be labeled by a domain expert, which will then be used as the labeled data set. However, existing approaches are computationally expensive and require searching through an entire unlabeled dataset, which may contain redundant instances that provide no instructive information to the classifier and can decrease the performance. To address this optimization problem, a hybrid system that combines active learning (AL) and particle swarm optimization (PSO) algorithms is proposed to reduce the cost of labeling while building a more efficient classifier. The novelty of this work resides in the integration of a bio-inspired optimization algorithm in the machine learning strategy. Furthermore, a novel uncertainty measure was integrated into the particle swarm optimization algorithm as an objective function to select from massive amounts of medical instances those that are deemed most informative. To evaluate the effectiveness of the proposed approach, eighteen (18) benchmark datasets were used and compared against three best-known classifiers with different learning paradigms: AL–NB an active learning algorithm using Naive Base classifier and Margin Sampling strategy, SVM (Support Vector Machine), ELM (Extreme Learning Machine) with supervised learning, and TSVM (Transductive Support Vector Machine) with the semi-supervised learning. Experiments showed that the proposed approach is effective in reducing the efforts required by experts for medical data annotation to produce an accurate classifier. The active learning approach has been utilized to optimize the expensive task of labeling. Based on a novel uncertainty measure, the nature-inspired algorithm PSO attempts to select from massive amounts of unlabeled medical instances those considered informative, at the same time improving the classifier performance. The experiments carried out confirm that the proposed strategy significantly enhances the performance of the AL algorithm compared with the commonly used uncertainty strategies. It achieves a performance similar to that of fully supervised and semi-supervised algorithms while requiring much less labeling. As a future extension of this work, it would be interesting to integrate other evolutionary optimization algorithms and compare them with our approach. In addition, it is beneficial to test the impact of using other variants of PSO algorithm in our approach. Also, it is aimed to test more classification algorithms in the experimentation process.

Special issue on SoC and AI processors

Special issue on SoC and AI processors

An interactive nonparametric evidential regression algorithm with instance selection

The nonparametric evidential regression (EVREG) method provides flexible forms of prediction regarding the value of output, allowing the output of training instances to be partially unknown. However, the superfluous training instances still have negative effects on the parameter learning in EVREG. To relax this limitation, this paper introduces an interactive nonparametric evidential regression (IEVREG) algorithm with instance selection. More specifically, the significance of an instance is firstly measured by defining the evaluation functions, taking into account both the prediction accuracy of regression model and the spatial information between that instance with other ones. According to a search strategy, the instances with high degree of significance are then selected to maximize an objective function. Different from existing instance selection methods, the selection of training instances is synchronously accomplished with the parameter learning in IEVREG, rather than just a separated data preprocessing operation as traditional methods do. Furthermore, the noise and redundant instances can be simultaneously removed and the performance of IEVREG is robust to the order of presentation of instances in raw data set. Experimental results show that the proposed IEVREG algorithm has appropriate prediction accuracy, while performing well selection of the representative training instances from the raw data set. Simulations on synthetic and UCI real-world data sets validate our conclusions.