UNSUPERVISED ENSEMBLE CLASSIFICATION FOR BIOMETRIC APPLICATIONS

2014 ◽  
Vol 28 (04) ◽  
pp. 1456007 ◽  
Author(s):  
ANA CERNEA ◽  
JUAN. LUIS. FERNÁNDEZ-MARTÍNEZ
Keyword(s):  
Face Recognition ◽  
Ensemble Learning ◽  
Nearest Neighbor ◽  
Ensemble Classification ◽  
Future Research ◽  
Ensemble Classifiers ◽  
Ensemble Techniques ◽  
Single Attribute ◽  
Consensus Classification ◽  
The Individual

In this paper, we propose different ensemble learning algorithms and their application to the face recognition problem. Three types of attributes are used for image representation: statistical, spectral, and segmentation features and regional descriptors. Classification is performed by nearest neighbor using different p-norms defined in the corresponding spaces of attributes. In this approach, each attribute together with its corresponding type of the analysis (local or global) and the distance criterion (norm or cosine), define a different classifier. The classification is unsupervised since no class information is used to improve the design of the different classifiers. Three different versions of ensemble classifiers are proposed in this paper: CAV1, CAV2, and CBAG, being the main differences among them the way the image candidates that perform the consensus are selected. The main results shown in this paper are the following: 1. The statistical attributes (local histogram and percentiles) are the individual classifiers that provided the higher accuracies, followed by the spectral methods (DWT), and the regional features (texture analysis). 2. No single attribute is able to provide systematically 100% accuracy over the ORL database. 3. The accuracy and stability of the classification is increased by consensus classification (ensemble learning techniques). 4. Optimum results are obtained by reducing the number of classifiers taking into account their diversity, and by optimizing the parameters of these classifiers using a member of the Particle Swarm Optimization (PSO) family. These results are in accord with the conclusions that are presented in the literature using ensemble learning methodologies, that is, it is possible to build strong classifiers by assembling different weak (or simple) classifiers based on different and diverse image attributes. Due to these encouraging results, future research will be devoted to the use of supervised ensemble techniques in face recognition and in other important biometric problems.

Exploring the Uncertainty Space of Ensemble Classifiers in Face Recognition

2015 ◽  
Vol 29 (03) ◽  
pp. 1556002 ◽  
Author(s):  
Juan Luis Fernández-Martínez ◽  
Ana Cernea
Keyword(s):  
Face Recognition ◽  
Nearest Neighbor ◽  
Learning Algorithm ◽  
Majority Voting ◽  
Borda Count ◽  
Discrete Wavelet ◽  
Final Decision ◽  
Ensemble Classifiers ◽  
Classification Problems ◽  
Uncertainty Space

In this paper, we present a supervised ensemble learning algorithm, called SCAV1, and its application to face recognition. This algorithm exploits the uncertainty space of the ensemble classifiers. Its design includes six different nearest-neighbor (NN) classifiers that are based on different and diverse image attributes: histogram, variogram, texture analysis, edges, bidimensional discrete wavelet transform and Zernike moments. In this approach each attribute, together with its corresponding type of the analysis (local or global), and the distance criterion (p-norm) induces a different individual NN classifier. The ensemble classifier SCAV1 depends on a set of parameters: the number of candidate images used by each individual method to perform the final classification and the individual weights given to each individual classifier. SCAV1 parameters are optimized/sampled using a supervised approach via the regressive particle swarm optimization algorithm (RR-PSO). The final classifier exploits the uncertainty space of SCAV1 and uses majority voting (Borda Count) as a final decision rule. We show the application of this algorithm to the ORL and PUT image databases, obtaining very high and stable accuracies (100% median accuracy and almost null interquartile range). In conclusion, exploring the uncertainty space of ensemble classifiers provides optimum results and seems to be the appropriate strategy to adopt for face recognition and other classification problems.

scholarly journals An Efficient Ensemble Learning Method for Gene Microarray Classification

2013 ◽  
Vol 2013 ◽  
pp. 1-10 ◽  
Author(s):  
Alireza Osareh ◽  
Bita Shadgar
Keyword(s):  
Feature Selection ◽  
Ensemble Learning ◽  
Feature Selection Method ◽  
Support Vector ◽  
Gene Microarray ◽  
Ensemble Classifiers ◽  
Classifier Ensembles ◽  
Rotation Forest ◽  
Ensemble Techniques ◽  
Effective Diagnosis

The gene microarray analysis and classification have demonstrated an effective way for the effective diagnosis of diseases and cancers. However, it has been also revealed that the basic classification techniques have intrinsic drawbacks in achieving accurate gene classification and cancer diagnosis. On the other hand, classifier ensembles have received increasing attention in various applications. Here, we address the gene classification issue using RotBoost ensemble methodology. This method is a combination of Rotation Forest and AdaBoost techniques which in turn preserve both desirable features of an ensemble architecture, that is, accuracy and diversity. To select a concise subset of informative genes, 5 different feature selection algorithms are considered. To assess the efficiency of the RotBoost, other nonensemble/ensemble techniques including Decision Trees, Support Vector Machines, Rotation Forest, AdaBoost, and Bagging are also deployed. Experimental results have revealed that the combination of the fast correlation-based feature selection method with ICA-based RotBoost ensemble is highly effective for gene classification. In fact, the proposed method can create ensemble classifiers which outperform not only the classifiers produced by the conventional machine learning but also the classifiers generated by two widely used conventional ensemble learning methods, that is, Bagging and AdaBoost.

scholarly journals Experimental Study and Comparison of Imbalance Ensemble Classifiers with Dynamic Selection Strategy

Entropy ◽  
2021 ◽  
Vol 23 (7) ◽  
pp. 822
Author(s):  
Dongxue Zhao ◽  
Xin Wang ◽  
Yashuang Mu ◽  
Lidong Wang
Keyword(s):  
Classification Performance ◽  
Ensemble Classification ◽  
Selection Strategy ◽  
Ensemble Classifiers ◽  
Imbalanced Datasets ◽  
Dynamic Selection ◽  
Imbalanced Classification ◽  
Ensemble Techniques ◽  
Imbalance Learning ◽  
Ensemble Algorithms

Imbalance ensemble classification is one of the most essential and practical strategies for improving decision performance in data analysis. There is a growing body of literature about ensemble techniques for imbalance learning in recent years, the various extensions of imbalanced classification methods were established from different points of view. The present study is initiated in an attempt to review the state-of-the-art ensemble classification algorithms for dealing with imbalanced datasets, offering a comprehensive analysis for incorporating the dynamic selection of base classifiers in classification. By conducting 14 existing ensemble algorithms incorporating a dynamic selection on 56 datasets, the experimental results reveal that the classical algorithm with a dynamic selection strategy deliver a practical way to improve the classification performance for both a binary class and multi-class imbalanced datasets. In addition, by combining patch learning with a dynamic selection ensemble classification, a patch-ensemble classification method is designed, which utilizes the misclassified samples to train patch classifiers for increasing the diversity of base classifiers. The experiments’ results indicate that the designed method has a certain potential for the performance of multi-class imbalanced classification.

Data-driven disruption prediction in GOLEM Tokamak using ensemble classifiers

2020 ◽  
Vol 39 (6) ◽  
pp. 8365-8376
Author(s):  
Jayakumar Chandrasekar ◽  
Surendar Madhawa ◽  
J. Sangeetha
Keyword(s):  
Predictive Model ◽  
Structural Damage ◽  
Machine Learning Algorithms ◽  
Data Driven ◽  
Ensemble Classifiers ◽  
Predictive Capacity ◽  
Ensemble Technique ◽  
Ensemble Techniques ◽  
The Individual ◽  
Base Learner

A robust disruption prediction system is mandatory in a Tokamak control system as the disruption can cause malfunctioning of the plasma-facing components and impair irrecoverable structural damage to the vessel. To mitigate the disruption, in this article, a data-driven based disruption predictor is developed using an ensemble technique. The ensemble algorithm classifies disruptive and non-disruptive discharges in the GOLEM Tokamak system. Ensemble classifiers combine the predictive capacity of several weak learners to produce a single predictive model and are utilized both in supervised and unsupervised learning. The resulting final model reduces the bias, minimizes variance and is unlikely to over-fit when compared to the individual model from a single algorithm. In this paper, popular ensemble techniques such as Bagging, Boosting, Voting, and Stacking are employed on the time-series Tokamak dataset, which consists of 117 normal and 70 disruptive shots. Stacking ensemble with REPTree (Reduced Error Pruning Tree) as a base learner and Multi-response Linear Regression as meta learner produced better results in comparison to other ensembles. A comparison with the widely employed stand-alone machine learning algorithms and ensemble algorithms are illustrated. The results show the excellent performance of the Stacking model with an F1 score of 0.973. The developed predictive model would be capable of warning the human operator with feedback about the feature(s) causing the disruption.

Development of a German-Language Questionnaire to Measure Collective Orientation as an Individual Attitude

2017 ◽  
Vol 76 (3) ◽  
pp. 91-105 ◽  
Author(s):  
Vera Hagemann
Keyword(s):  
Team Performance ◽  
Reliability And Validity ◽  
Future Research ◽  
German Language ◽  
Team Members ◽  
Collective Orientation ◽  
Team Research ◽  
German Speaking ◽  
Individual Variables ◽  
The Individual

Abstract. The individual attitudes of every single team member are important for team performance. Studies show that each team member’s collective orientation – that is, propensity to work in a collective manner in team settings – enhances the team’s interdependent teamwork. In the German-speaking countries, there was previously no instrument to measure collective orientation. So, I developed and validated a German-language instrument to measure collective orientation. In three studies (N = 1028), I tested the validity of the instrument in terms of its internal structure and relationships with other variables. The results confirm the reliability and validity of the instrument. The instrument also predicts team performance in terms of interdependent teamwork. I discuss differences in established individual variables in team research and the role of collective orientation in teams. In future research, the instrument can be applied to diagnose teamwork deficiencies and evaluate interventions for developing team members’ collective orientation.

Academic Events: An Empirically Grounded Typology and Their Academic Impact

2020 ◽  
Vol 24 (4) ◽  
pp. 481-497 ◽  
Author(s):  
Thomas Trøst Hansen ◽  
David Budtz Pedersen ◽  
Carmel Foley
Keyword(s):  
Analytical Framework ◽  
Future Research ◽  
Academic Impact ◽  
Future Research Agenda ◽  
Tourism Studies ◽  
Academic Events ◽  
Business Events ◽  
The Individual ◽  
Analytical Frameworks ◽  
Academic Sector

The meetings industry, government bodies, and scholars within tourism studies have identified the need to understand the broader impact of business events. To succeed in this endeavor, we consider it necessary to develop analytical frameworks that are sensitive to the particularities of the analyzed event, sector, and stakeholder group. In this article we focus on the academic sector and offer two connected analyses. First is an empirically grounded typology of academic events. We identify four differentiating dimensions of academic events: size, academic focus, participants, and tradition, and based on these dimensions we develop a typology of academic events that includes: congress, specialty conference, symposium, and practitioners' meeting. Secondly, we outline the academic impact of attending these four types of events. For this purpose, the concept of credibility cycles is used as an analytical framework for examining academic impact. We suggest that academic events should be conceptualized and evaluated as open marketplaces that facilitate conversion of credibility. Data were obtained from interviews with 22 researchers at three Danish universities. The study concludes that there are significant differences between the events in terms of their academic impact. Moreover, the outcome for the individual scholar depends on the investment being made. Finally, the study calls for a future research agenda on beyond tourism benefits based on interdisciplinary collaborations.

Physiology of Larval Feeding

2018 ◽  
Keyword(s):  
Growth And Development ◽  
Marine Invertebrate ◽  
Larval Growth ◽  
Circulatory System ◽  
Future Research ◽  
Larval Feeding ◽  
Digestive Physiology ◽  
Larval Body ◽  
The Individual ◽  
Invertebrate Larvae

The functional properties of marine invertebrate larvae represent the sum of the physiological activities of the individual, the interdependence among cells making up the whole, and the correct positioning of cells within the larval body. This chapter examines physiological aspects of nutrient acquisition, digestion, assimilation, and distribution within invertebrate larvae from an organismic and comparative perspective. Growth and development of larvae obviously require the acquisition of “food.” Yet the mechanisms where particulate or dissolved organic materials are converted into biomass and promote development of larvae differ and are variably known among groups. Differences in the physiology of the digestive system (secreted enzymes, gut transit time, and assimilation) within and among feeding larvae suggest the possibility of an underappreciated plasticity of digestive physiology. How the ingestion of seawater by and the existence of a circulatory system within larvae contribute to larval growth and development represent important topics for future research.

The Role of Normative Information in Judgments of Others

2019 ◽  
Author(s):  
Katherine H. Rogers
Keyword(s):  
Social Desirability ◽  
Interpersonal Perception ◽  
Future Research ◽  
Average Person ◽  
The Individual ◽  
The Relationship ◽  
Source Of Information

When forming impressions of an other’s personality, people often rely on information not directly related to the individual at hand. One source of information that can influence people’s impressions of others is the personality of the average person (i.e., normative profile). This relationship between the normative profile and an impression is called normative accuracy or normativity. In this chapter, you will learn about the average personality, why it is important, the relationship to social desirability and what it means to have a normative impression, as well as correlates and moderators of normativity. More broadly, you will learn about current research and views regarding the normative profile and normative impressions as well as concrete steps for incorporating this approach into your future research on interpersonal perception.

scholarly journals Opportunities and Challenges of Green-Lean: An Integrated System for Sustainable Construction

2020 ◽  
Vol 12 (11) ◽  
pp. 4460 ◽  
Author(s):  
Mohammadsoroush Tafazzoli ◽  
Ehsan Mousavi ◽  
Sharareh Kermanshachi
Keyword(s):  
Cost Savings ◽  
Integrated Approach ◽  
Integrated System ◽  
Waste Minimization ◽  
Sustainable Construction ◽  
Future Research ◽  
Safety Improvement ◽  
The Individual ◽  
High Level ◽  
Industry Professionals

Although the two concepts of lean and sustainable construction have been developed due to different incentives, and they do not pursue the same exact goals, there exists considerable commonality between them. This paper discusses the potentials for integrating the two approaches and their practices and how the resulting synergy from combining the two methods can potentially lead to higher levels of fulfilling the individual goals of each of them. Some limitations and challenges to implementing the integrated approach are also discussed. Based on a comprehensive review of existing papers related to sustainable and lean construction topics, the commonality between the two approaches is discussed and grouped in five categories of (1) cost savings, (2) waste minimization, (3) Jobsite safety improvement, (4) reduced energy consumption, and (5) customers’ satisfaction improvement. The challenges of this integration are similarly identified and discussed in the four main categories of (1) additional initial costs to the project, (2) difficulty of providing specialized expertise, (3) contractors’ unwillingness to adopt the additional requirements, and (4) challenges to establish a high level of teamwork. Industry professionals were then interviewed to rank the elements in each of the two categories of opportunities and challenges. The results of the study highlight how future research can pursue the development of a new Green-Lean approach by investing in the communalities and meeting the challenges of this integration.

scholarly journals Validity of the Gender Dysphoria diagnosis and incidence trends in Sweden: a nationwide register study

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Malin Indremo ◽  
Richard White ◽  
Thomas Frisell ◽  
Sven Cnattingius ◽  
Alkistis Skalkidou ◽  
...  
Keyword(s):  
Gender Dysphoria ◽  
Age Groups ◽  
Medical Intervention ◽  
Future Research ◽  
Advantages And Disadvantages ◽  
Incidence Trends ◽  
Interaction Terms ◽  
Definition Of ◽  
The Individual ◽  
Nationwide Register Study

AbstractThe aim of this study was to examine the validity of the Gender Dysphoria (GD) diagnoses in the Swedish National Patient Register (NPR), to discuss different register-based definitions of GD and to investigate incidence trends. We collected data on all individuals with registered GD diagnoses between 2001 and 2016 as well as data on the coverage in the NPR. We regarded gender confirming medical intervention (GCMI) as one proxy for a clinically valid diagnosis and calculated the positive predictive value (PPV) for receiving GCMI for increasing number of registered GD diagnoses. We assessed crude and coverage-adjusted time trends of GD during 2004–2015 with a Poisson regression, using assigned sex and age as interaction terms. The PPV for receiving GCMI was 68% for ≥ 1 and 79% for ≥ 4 GD-diagnoses. The incidence of GD was on average 35% higher with the definition of ≥ 1 compared to the definition of ≥ 4 diagnoses. The incidence of GD, defined as ≥ 4 diagnoses increased significantly during the study period and mostly in the age categories 10–17 and 18–30 years, even after adjusting for register coverage. We concluded that the validity of a single ICD code denoting clinical GD in the Swedish NPR can be questioned. For future research, we propose to carefully weight the advantages and disadvantages of different register-based definitions according to the individual study’s needs, the time periods involved and the age-groups under study.

Sign in / Sign up

Export Citation Format

Share Document