Irrelevant Ones Research Articles

Task difficulty modulates voluntary attention allocation, but not distraction in an auditory distraction paradigm

Keeping task-relevant sensory events in the focus of attention while ignoring irrelevant ones is crucial for optimizing task behavior. This attention-distraction balance might change with the perceptual demands of the ongoing task: while easy tasks might be performed with low attentional effort, difficult ones require enhanced attention. The goal of the present study was to investigate how task difficulty affected allocation of attention and distractibility in an auditory distraction paradigm. Participants performed a tone duration discrimination task in which tones were rarely, occasionally presented at a rare pitch (distracters), and task difficulty was manipulated by the duration difference between short and long tones. Short tones were consistently 200 ms long, while long tone duration was 400 ms in the easy, and 260 ms in the difficult condition. Behavioral results and deviant-minus-standard event-related potential (ERP) waveforms suggested similar magnitudes of distraction in both conditions. ERPs without such a subtraction showed that tone onsets were preceded by a negative-going trend, suggesting that participants prepared for tone onsets. In the difficult condition, N1 amplitudes to tone onsets were enhanced, indicating that participants invested more attentional resources. Increased difficulty also slowed down tone offset processing as reflected by significantly delayed offset-related P1 and N1/N2 waveforms. These results suggest that although task difficulty compels participants to attend the tones more strongly, this does not have significant impact on distraction-related processing.

Unsupervised satellite image classification based on partial transfer learning

Satellite image classification plays an important role in many fields. It can be divided into two groups: supervised and unsupervised classification. The former requires a large number of labeled data, while, in practice, satellite images usually lack of sufficient ones. How to achieve high accuracy on unsupervised satellite image classification is a key problem. For tackling this problem, based on the idea of partial transfer learning, we propose an novel end-to-end unsupervised classification method called coordinate partial adversarial domain adaptation (CPADA) for satellite image classification. Under the aid of a novel coordinate loss, our framework transfers relevant examples in the shared classes to promote performance, and ignore irrelevant ones in the specific classes to mitigate negative transfer. Experiments show that our CPADA exceeds state-of-the-art results for unsupervised satellite image classification task.

Emergency medical services in Ethiopia: Drivers, challenges and opportunities.

Ethiopia has a well established health care system but lacks significant improvements on emergency medical services and suffers a shortage of equally initiative among all regional states and city administration of Ethiopia. This study aimed to examine the drivers, challenges, and opportunities of Emergency Medical Services (EMS) and to identify new evidence for future policy making in Ethiopia. A narrative review of the literature related to EMS was undertaken to describe the drivers, challenges, and opportunities for EMS in Ethiopia from July, 2000 to September, 2018. The search was done from four relevant electronic databases: MEDLINE, Science Directs, Scopus and PubMed by using Google Scholar and Google with key search words used mainly as "Emergency Medical services in Ethiopia". The inclusion criteria were an original study or review studies involving Emergency Medical Services in Ethiopia. Among the available papers, the relevant articles were selected while the irrelevant ones were excluded. There was lack of trained emergency medical providers and misdistribution of trained professionals, immaturity of the program, lack of partnership and stakeholders and lack of motivation towards Emergency medical services. Emergency medical services hamper significant problems similar to other African countries that required being addressed in Ethiopia context for achieving the program and in order to obtain intended outcomes for the country. A long-term discussion is needed to further improve the services system in various health care facilities. An Emergency Medical services policy making and analysis framework is needed to make quality emergency medical care at Emergency department in hospitals and outside the hospitals.

Enhance code search via reformulating queries with evolving contexts

To improve code search, many query expansion (QE) approaches use APIs or crowd knowledge for expanding a query. However, these approaches may sometimes negatively impact the retrieval performance. This is because they can’t distinguish the relevant terms from the irrelevant ones among a large set of candidate expansion terms and expand a query with irrelevant terms. In this paper, we propose QREC, a query reformulation approach with evolving contexts that refer to new/deleted terms and dependent terms during the code evolution. By considering the new terms as the relevant and the deleted terms as the irrelevant, QREC could reformulate a query with appropriate expansion terms. The experimental results show that QREC outperforms the state-of-the-art QE approaches (e.g., CodeHow and QECK) by 9–11% and improves the precision of the code search algorithms IR, Portfolio and VF by up to 37–45%.

Access to consciousness of briefly presented visual events is modulated by transcranial direct current stimulation of left dorsolateral prefrontal cortex

Adaptive behaviour requires the ability to process goal-relevant events at the expense of irrelevant ones. However, perception of a relevant visual event can transiently preclude access to consciousness of subsequent events — a phenomenon called attentional blink (AB). Here we investigated involvement of the left dorsolateral prefrontal cortex (DLPFC) in conscious access, by using transcranial direct current stimulation (tDCS) to potentiate or reduce neural excitability in the context of an AB task. In a sham-controlled experimental design, we applied between groups anodal or cathodal tDCS over the left DLPFC, and examined whether this stimulation modulated the proportion of stimuli that were consciously reported during the AB period. We found that tDCS over the left DLPFC affected the proportion of consciously perceived target stimuli. Moreover, anodal and cathodal tDCS had opposing effects, and exhibited different temporal patterns. Anodal stimulation attenuated the AB, enhancing conscious report earlier in the AB period. Cathodal stimulation accentuated the AB, reducing conscious report later in the AB period. These findings support the notion that the DLPFC plays a role in facilitating information transition from the unconscious to the conscious stage of processing.

Memory-Augmented Temporal Dynamic Learning for Action Recognition

Human actions captured in video sequences contain two crucial factors for action recognition, i.e., visual appearance and motion dynamics. To model these two aspects, Convolutional and Recurrent Neural Networks (CNNs and RNNs) are adopted in most existing successful methods for recognizing actions. However, CNN based methods are limited in modeling long-term motion dynamics. RNNs are able to learn temporal motion dynamics but lack effective ways to tackle unsteady dynamics in long-duration motion. In this work, we propose a memory-augmented temporal dynamic learning network, which learns to write the most evident information into an external memory module and ignore irrelevant ones. In particular, we present a differential memory controller to make a discrete decision on whether the external memory module should be updated with current feature. The discrete memory controller takes in the memory history, context embedding and current feature as inputs and controls information flow into the external memory module. Additionally, we train this discrete memory controller using straight-through estimator. We evaluate this end-to-end system on benchmark datasets (UCF101 and HMDB51) of human action recognition. The experimental results show consistent improvements on both datasets over prior works and our baselines.

The effect of verbal information about model’s competence on overimitation by preschool children

Overimitation is defined by a tendency to copy all actions executed by a model, even the clearly irrelevant ones. The motivational mechanisms and functionality of overimitation are still not well understood, but its possible adaptive meaning could be related to causal opacity of a great part of socially learned behaviors. This phenomenon has been widely replicated in several contexts and has been observed in the behavior of children over 2 years of age and even in adults. Despite the seeming robustness of overimitation, studies have shown that it is sensitive to some characteristics of a model observed such as age, familiarity, proficiency, and reputation. Our work intended to investigate the effect of information about the competence of an adult model on the copying of irrelevant actions by preschool children (5 years old) in a task. We tested the influence of self-declared information about the model competence and of the same kind of information given by third parties in a conversation about the model. Our results reveal no effect of both “self-declared competence” and “reputation” biases on overimitation. We discuss that this result may have occurred because other information available to participants, and not manipulated by us, was used to infer model competence such as the model’s age and success in the task directly observed by the participants. Another potential explanation is that children use a “copy all, correct later” strategy in a context where only one model is available.

4D facial expression recognition using multimodal time series analysis of geometric landmark-based deformations

One of the main challenges in dynamic facial expression recognition is how to capture temporal information in the system. In this study, a novel approach based on time series analysis is adapted for this problem. The proposed dynamic facial expression recognition system comprises four phases: head pose correction and normalization, feature extraction, feature selection and classification. Head pose detection and correction is the first phase to realign locations of the facial landmarks. A comprehensive set of geometric deformations including point, distance and angle deformations are extracted from the key points. The concept of facial action unit analysis is interlocked with this phase to identify related key points from the landmarks. A set of multimodal time series are then constructed from the extracted deformations by applying a sliding window to characterize the dynamics of mean deformations in a window. In the third phase, a feature selection method based on neighborhood component analysis is applied on the peak value of deformation features to select useful features and discard irrelevant ones. Finally, adaptive cost dynamic time warping is utilized to recognize six prototypic expressions from multimodal time series of selected features. Experimental results on BU-4DFE data set confirm that proposed algorithm is efficient in dynamic facial expression recognition compared with state of the art.

Evolutionary feature subsets selection based on interaction information for high dimensional imbalanced data classification

With the advent of technology in various scientific fields, high dimensional data are becoming abundant. A general approach to tackle the resulting challenges is to reduce data dimensionality through feature selection. Traditional feature selection approaches concentrate on selecting relevant features and ignoring irrelevant or redundant ones. However, most of these approaches neglect feature interactions. On the other hand, some datasets have imbalanced classes, which may result in biases towards the majority class. The main goal of this paper is to propose a novel feature selection method based on the interaction information (II) to provide higher level interaction analysis and improve the search procedure in the feature space. In this regard, an evolutionary feature subset selection algorithm based on interaction information is proposed, which consists of three stages. At the first stage, candidate features and candidate feature pairs are identified using traditional feature weighting approaches such as symmetric uncertainty (SU) and bivariate interaction information. In the second phase, candidate feature subsets are formed and evaluated using multivariate interaction information. Finally, the best candidate feature subsets are selected using dominant/dominated relationships. The proposed algorithm is compared with some other feature selection algorithms including mRMR, WJMI, IWFS, IGFS, DCSF, IWFS, K_OFSD, WFLNS, Information Gain and ReliefF in terms of the number of selected features, classification accuracy, F-measure and algorithm stability using three different classifiers, namely KNN, NB, and CART. The results justify the improvement of classification accuracy and the robustness of the proposed method in comparison with the other approaches.

A fuzzy agent approach for smart data extraction in big data environments

The era of big data has brought new challenges in data processing ad management. Existing analytical tools are now close to facing ongoing challenges thus providing satisfactory results at a reasonable cost. However, the velocity at which new data are flooded and the noise generated from such a large volume leads to various new challenges.The present research combines two artificial intelligence fields the represented by multi-agent technologies and fuzzy logic inference systems in order to extract the needed smart data from big noisy ones. A multi-fuzzy agent-based large-scale wireless sensor network has been used to demonstrate the effectiveness of the proposed approach. It handles sensors as autonomous fuzzy agents to measure the relevance of the collected data and eliminate the irrelevant ones. The results of the simulation exhibit a high quality of the data with a decrease in the sensors energy consumption, leading to a longer lifetime of the network.

Efficient Detection of Buried Plastic Pipes by Combining GPR and Electric Field Methods

In this paper, an efficient plastic pipe detecting model is proposed, which combines the ground penetrating radar (GPR) and the electric field method. The model consists of the electric field locating model (EFLM) and the GPR B-scan image interpreting (GBII) model. Synchronized electric field and GPR data are collected through a data acquisition device dedicatedly designed for the swift and accurate estimation of buried plastic pipes. The EFLM estimates the approximate locations of underground plastic pipes from the electric field data quickly, separates a GPR B-scan image into segments, keeps the segments that might contain hyperbolas, and discards the irrelevant ones. Then, the GBII model interprets the depth and radius of the buried pipe in the kept segments. Our numerical simulations and experiments prove that by utilizing the EFLM, the 1-D electric field data could be processed quickly and the GPR B-scan image could be segmented with part of irrelevant pixels discarded, while hyperbolas in the kept image segments could be automatically and accurately fitted. With our proposed model, the depth and radius of the buried pipes could be efficiently obtained.

Stopping rules for mutual information-based feature selection

In recent years feature selection methods based on mutual information have attracted a significant attention. Most of the proposed methods are based on sequential forward search which add at each step a feature which is most relevant in explaining a class variable when considered together with the already chosen features. Such procedures produce ranking of features ordered according to their relevance. However significant limitation of all existing methods is lack of stopping rules which separate relevant features placed on the top of the ranking list from irrelevant ones. Finding an appropriate stopping rule is particularly important in domains where one wants to precisely determine the set of features affecting the class variable and discard the irrelevant ones (e.g. in genome-wide association studies the goal is to precisely determine mutations in DNA affecting the disease). In this work we propose stopping rules which are based on distribution of approximation of conditional mutual information given that all relevant features have been already selected. We show that the distribution is approximately chi square with appropriate number of degrees of freedom provided features are discretized into moderate number of bins. The proposed stopping rules are based on quantiles of the distribution and related p-values which are compared with thresholds used in multiple hypothesis testing. Importantly the proposed methods do not require additional validation data and are independent from the classifier. The extensive simulation experiments indicate that the rules separate relevant features from the irrelevant ones. We show experimentally that Positive Selection Rate (fraction of features correctly selected as relevant with respect to all relevant features) approaches 1, when sample size increases. At the same time, False Discovery Rate (fraction of irrelevant features selected with respect to all selected features) is controlled. The experiments on 17 benchmark datasets indicate that the classification models, built on features selected by the proposed methods, in 13 cases achieve significantly higher accuracy than the models based on all available features.

Stepwise extreme learning machine for statistical downscaling of daily maximum and minimum temperature

Increasing temperature from climate change can bring a number of different risks such as more droughts and heat waves, and increasing sea level rise. Assessment of climate change with future scenarios is essential to adapt these impacts. To provide climate change information through the outputs of general circulation models at finer resolution, a reliable and accurate downscaling model has always been of great interest. Meeting this need, artificial neural network (ANN) has been commonly employed in downscaling for nonlinear models. Extreme learning machine (ELM), a recently developed ANN, is an efficient learning algorithm for generalized single hidden layer feedforward neural networks. In light of its simple learning algorithm, we introduced a useful approach to combine the stepwise feature selection method into ELM for temperature downscaling, as stepwise ELM (SWELM), since model complexity and computational time consumption of a traditional ANN impedes application of stepwise feature selection. This SWELM is able to identify the most influential predictors in a dataset and use them to train a nonlinear model while removing the irrelevant ones. The ELM and SWELM as well as regular ANN were tested in a simulation study. Results indicated that ELM even with randomness of weights and biases in the nodes of input and hidden layers better performed than did ANN. Also, SWELM presents a capability to select the influential predictors and remove the unrelated variables. A case study with downscaling temperature of Wisconsin, USA, showed that ELM was a comparable alternative to ANN. SWELM outperformed the ANN algorithm for temperature downscaling and sometimes predicted the temperature increase larger than did others for future scenarios. The current study of temperature downscaling with the statistical tool allows assessing the possible impacts of climate change in a local scale and some developing countries where sophisticate research cannot be eligible.

IMPULSE BUYING BEHAVIOR: A STUDY ON GROCERY ITEMS

The spontaneous or unplanned purchase of goods is the best way to define impulse buying. Impulse buying can happen in any product ranging from new product, samples to well establish products with va rious discount offers. There is limited work had been conducted to assess the impulse buying habit of customer in FMCG category. The formulated hypothesis is verified by using a selfreported questionnaire. The questionnaire contains the demographic information of respondents and 16 statements. A survey was planned and conducted on convenience basis. During survey 215 grocery items customers are approached and finally the sample size of the study is 179 by dropping the irrelevant ones. Impulsive buying is very important for grocery items. Fast Moving Consumer Goods (FMCG) companies and specially grocery item company should take into account what factors motivates and what factors de-motivates impulse buying, product category in grocery item which leads to maximum impulse buying and understand relationship between demographic detail and impulse buying in grocery items.

Feature selection of non-intrusive load monitoring system using RFE and RF

Feature selection is a process of selecting relevant features and removing the irrelevant and redundant ones based on certain specific criteria in the original date set. This paper proposes the random forest algorithm (RF) based on recursive feature elimination (RFE) to select features in non-intrusive load monitoring system. Adopting the random forest algorithm as the basic approach, we repeatedly build models and filter features to identify the best subset of features. Last, the results of experiments using public data sets verify the effectiveness of the proposed algorithm.

Effect of Principal Component Analysis in Feature based Uncalibrated Steganalysis using Block Dependency

Steganalysis is a domain where hidden information sent through a medium over the internet is detected. The medium can be text, audio, images or video. The steganalytic method is of two types- target and blind. In this paper, we utilize the blind steganalysis and extract the statistical changes that occur when an image is embedded. The statistical values are known as features. During the extraction of features, the irrelevant ones are also extracted, which may hamper the efficiency of the analysis. This paper checks the efficiency rate of steganalysis by eliminating irrelevant data. Principal component analysis is used for feature reduction. Discrete Cosine Transform is used. The features used here are a combination of different statistical features as first order, second order, and Markov. The features which include the inter-block dependency features, as well as the intra-block features are reduced.

A hybrid model of fuzzy min–max and brain storm optimization for feature selection and data classification

Swarm intelligence (SI)-based optimization methods have been extensively used to tackle feature selection problems. A feature selection method extracts the most significant features and removes irrelevant ones from the data set, in order to reduce feature dimensionality and improve the classification accuracy. This paper combines the incremental learning Fuzzy Min–Max (FMM) neural network and Brain Storm Optimization (BSO) to undertake feature selection and classification problems. Firstly, FMM is used to create a number of hyperboxes incrementally. BSO, which is inspired by the human brainstorming process, is then employed to search for an optimal feature subset. Ten benchmark problems and a real-world case study are conducted to evaluate the effectiveness of the proposed FMM-BSO. In addition, the bootstrap method with the 95% confidence intervals is used to quantify the results statistically. The experimental results indicate that FMM-BSO is able to produce promising results as compared with those from the original FMM network and other state-of-the-art feature selection methods such as particle swarm optimization, genetic algorithm, and ant lion optimization.

Assessing variable importance in clustering: a new method based on unsupervised binary decision trees

We consider different approaches for assessing variable importance in clustering. We focus on clustering using binary decision trees (CUBT), which is a non-parametric top-down hierarchical clustering method designed for both continuous and nominal data. We suggest a measure of variable importance for this method similar to the one used in Breiman’s classification and regression trees. This score is useful to rank the variables in a dataset, to determine which variables are the most important or to detect the irrelevant ones. We analyze both stability and efficiency of this score on different data simulation models in the presence of noise, and compare it to other classical variable importance measures. Our experiments show that variable importance based on CUBT is much more efficient than other approaches in a large variety of situations.

SP-Net: A Novel Framework to Identify Composite Sketch

Identifying composite sketches with digital face photos is an important and challenging task for law enforcement agencies. It has attracted a wide research interest in the face recognition area. In this paper, we present a novel framework that identifies the photo corresponding to a given composite face sketch. A coupled deep convolutional neural network, named Sketch-Photo Net (SP-Net) is proposed, which is fed with a positive or negative photo-sketch pair. In the proposed SP-Net, the customized VGG-Face network is adopted as base model and is followed by two branches, namely S-Net and P-Net, for sketch and photo, respectively. The S-Net and the P-Net are able to learn discriminative features between the sketches and the photos, regardless of the appearance gap by introducing the concept of elastic learning. In other words, to extract the most important features from the input, the network needs to learn the relevant features along with the irrelevant ones. To do so, higher dimension layers are used after the three 512 layers from VGG-FaceNet. Since the network learns representative features, we decrease the dimension of the layers to produce the most representative features. In addition, contrastive loss is employed to discover the coherent visual structures between sketch and photo. Experimental results on E-PRIP face sketch dataset indicate that the proposed network significantly outperforms the state-of-the-art composite sketch identification methods.

Variable selection via adaptive false negative control in linear regression

Variable selection methods have been developed in linear regression to provide sparse solutions. Recent studies have focused on further interpretations on the sparse solutions in terms of false positive control. In this paper, we consider false negative control for variable selection with the goal to efficiently select a high proportion of relevant predictors. Different from existing studies in power analysis and sure screening, we propose to directly estimate the false negative proportion (FNP) of a decision rule and select the smallest subset of predictors that has the estimated FNP less than a user-specified control level. The proposed method is adaptive to the user-specified control level on FNP by selecting less candidates if a higher level is implemented. On the other hand, when data has stronger effect size or larger sample size, the proposed method controls FNP more efficiently with less false positives. New analytic techniques are developed to cope with the major challenge of FNP control when relevant predictors cannot be consistently separated from irrelevant ones. Our numerical results are in line with the theoretical findings.