Using Mahalanobis distance and decision tree to analyze abnormal patterns of behavior in a maintenance outsourcing process-a case study

2020 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Ssu-Han Chen ◽  
Yiyo Kuo ◽  
Jin-Kwan Lin
Keyword(s):  
Decision Tree ◽  
Behavior Pattern ◽  
Abnormal Behavior ◽  
Behavior Patterns ◽  
Minority Class ◽  
Content Type ◽  
Empirical Results ◽  
Integrated Methodology ◽  
Maintenance Outsourcing ◽  
Patterns Of Behavior

PurposeThe purpose of this paper is to analyze abnormal behavior patterns in a maintenance outsourcing process. Based on the results, the managers can focus on the abnormal behavior and the direction of the investigation can be narrowed. The abnormal behavior can be identified more easily.Design/methodology/approachMaholanobis Distance (MD) and Decision Tree (DT) are integrated to analyze for abnormal behavior patterns. To prevent abnormal behaviors, a maintenance outsourcing case must be passed by several managers in different departments. In this research, some criteria for pairs of managers are calculated first. Based on the criteria, the MDs of these pairs can be calculated. Pairs are categorized by their MDs. Any pair whose MD is higher than a threshold is labeled “abnormal” while the remaining are labeled “normal”. After oversampling the minority class of abnormal, a DT is built by Classification and Regression Trees (CART) based on the labeled dataset. Finally, the combination of criteria for abnormal categories is extracted from the tree.FindingsThrough the results from the DT, the combinations of criteria provide obvious characteristics of cases that are categorized as abnormal, and then provide a direction for investigators. Thus, the range of investigation can be narrowed. The empirical results show that the result of the proposed integrated methodology is helpful for abnormal behavior pattern analysis.Practical implicationsThis research is intended to help an organization to enhance their investigation in a large number of maintenance outsourcing cases. About 8,000 cases are collected for analysis.Originality/valueThe integration of MD and DT for analyzing abnormal behavior patterns in a maintenance outsourcing process is not found in the literature. Moreover, the empirical results show that the proposed integrated methodology is helpful in a real application.

Identification of substructures in complex networks using formal concept analysis

2018 ◽  
Vol 14 (3) ◽  
pp. 281-298 ◽  
Author(s):  
Sebastião M. Neto ◽  
Sérgio Dias ◽  
Rokia Missaoui ◽  
Luis Zárate ◽  
Mark Song
Keyword(s):  
Social Networks ◽  
Social Network ◽  
Formal Concept Analysis ◽  
Computational Models ◽  
Behavior Pattern ◽  
Concept Analysis ◽  
Behavior Patterns ◽  
Formal Concept ◽  
Content Type ◽  
Access Patterns

Purpose In recent years, the increasing complexity of the hyper-connected world demands new approaches for social network analysis. The main challenges are to find new computational methods that allow the representation, characterization and analysis of these social networks. Nowadays, formal concept analysis (FCA) is considered an alternative to identifying conceptual structures in a social network. In this FCA-based work, this paper aims to show the potential of building computational models based on implications to represent and analyze two-mode networks. Design/methodology/approach This study proposes an approach to find three important substructures in social networks such as conservative access patterns, minimum behavior patterns and canonical access patterns. The present study approach considered as a case study a database containing the access logs of a cable internet service provider. Findings The result allows us to uncover access patterns, conservative access patterns and minimum access behavior patterns. Furthermore, through the use of implications sets, the relationships between event-type elements (websites) in two-mode networks are analyzed. This paper discusses, in a generic form, the adopted procedures that can be extended to other social networks. Originality/value A new approach is proposed for the identification of conservative behavior in two-mode networks. The proper implications needed to handle minimum behavior pattern in two-mode networks is also proposed to be analyzed. The one-item conclusion implications are easy to understand and can be more relevant to anyone looking for one particular website access pattern. Finally, a method for a canonical behavior representation in two-mode networks using a canonical set of implications (steam base), which present a minimal set of implications without loss of information, is proposed.

An investigation of trust in e-banking

2015 ◽  
Vol 38 (12) ◽  
pp. 1267-1284 ◽  
Author(s):  
Chian-Son Yu ◽  
Mehdi Asgarkhani
Keyword(s):  
Significant Influence ◽  
Holistic View ◽  
Banking Services ◽  
Content Type ◽  
Empirical Results ◽  
Other Information ◽  
Information And Communication ◽  
Structural Assurance ◽  
Antecedents And Consequences ◽  
Academic Implications

Purpose – The purpose of this paper is to understand the connection among trust’s antecedents, dimensions and consequences in the context of e-banking. Design/methodology/approach – A survey of 510 and 122 respondents in Taiwan and New Zealand (NZ), respectively, was conducted. Findings – The empirical results indicate that, first, not all trusts’ precursors the authors considered have significant influence on generating consumers’ trust and, second, that influential weights of these precursors on building consumer trust vary across consumers and cultures. Meanwhile, all factors on the e-banking side hold greatly significant influence on consumers’ trust in both NZ and Taiwan cases. Research limitations/implications – Practical and academic implications culled from the empirical results are discussed, and these implications may also be applicable to other information and communication technology (ICT) solutions and innovation banking services. Practical implications – Before banks shift their focus on to trust resources of consumer side, banks are advised to create clients’ trust from e-banking side, such as situational normality and structural assurance. Originality/value – This paper takes a holistic view to investigate the links between trust’s dimensions, antecedents and consequences in a single research structure, and the implications may also be applicable to other ICT solutions and innovative banking services.

P1-559: TOWARD PERCEPTUAL DEEP LEARNING-BASED ABNORMAL BEHAVIOR PATTERN ANALYSIS FOR PATIENTS WITH ALZHEIMER'S DISEASE

2006 ◽  
Vol 14 (7S_Part_10) ◽  
pp. P548-P548
Author(s):  
Jeonghwan Gwak ◽  
Jong-In Song ◽  
Kiseon Kim ◽  
Moongu Jeon ◽  
Cheolbin Park ◽  
...  
Keyword(s):  
Alzheimer’S Disease ◽  
Alzheimer's Disease ◽  
Deep Learning ◽  
Pattern Analysis ◽  
Behavior Pattern ◽  
Abnormal Behavior

Stakeholder dialogue on Facebook

2017 ◽  
Vol 11 (2) ◽  
pp. 257-267 ◽  
Author(s):  
Sebastian Martin
Keyword(s):  
Design Methodology ◽  
Online Survey ◽  
Energy Sector ◽  
Public Ownership ◽  
Stakeholder Dialogue ◽  
Dialogic Communication ◽  
Content Type ◽  
Empirical Results ◽  
Utility Companies

Purpose German, Austrian and Swiss utilities are confronted with radical changes in the European energy sector. A dialogue between the utility companies and their various groups of stakeholders is gaining importance. Increasingly, utilities create their own Facebook presence enabling such a dialogue. Still, to the best of the author’s knowledge there exists no research which explicitly focuses the stakeholder dialogue of German, Austrian or Swiss utilities on Facebook. Therefore, the purpose of this paper is to analyse Facebook as an instrument for dialogic communication in the energy sector. Design/methodology/approach An online survey was distributed to 1,280 German, Austrian and Swiss utilities, and 14 per cent of the utilities completed the survey, including 130 German, 19 Austrian and 25 Swiss companies. The participating utilities are primarily in public ownership. Findings The Facebook conversation of utility companies and their stakeholders meets the basic requirements of a virtual stakeholder dialogue. Nevertheless, less than half of the companies perceive their current stakeholder conversation on Facebook as truly interactive. Therefore, even if the basic requirements of a dialogue are met, most companies still do not seem to fully use the dialogue potential of Facebook. Originality/value This study provides first insights into virtual stakeholder dialogues in the energy sector. A suggestion to operationalise such a virtual dialogue is provided. Both operationalisation as well as the empirical results help researchers and practitioners to better understand virtual stakeholder dialogues.

The influence of culture in entrepreneurs’ opportunity exploitation decision in Tanzania

2019 ◽  
Vol 11 (1) ◽  
pp. 22-43 ◽  
Author(s):  
Jia Liu ◽  
Frida Thomas Pacho ◽  
Wang Xuhui
Keyword(s):  
Risk Taking ◽  
Power Distance ◽  
Uncertainty Avoidance ◽  
Formal Institutions ◽  
Entrepreneurial Opportunity ◽  
Content Type ◽  
Opportunity Exploitation ◽  
Empirical Results ◽  
Two Cities ◽  
Risk Taking Behavior

Purpose The purpose of this paper is to empirically explore the impact of culture (using individualism, power distance and uncertainty avoidance) on entrepreneurial risk taking behavior which leads to the opportunity exploitation decision. Moreover, it also uses risk taking behavior of entrepreneurial as the mediation variable between culture and opportunity exploitations decisions. Design/methodology/approach The study took place in Tanzania, which is allocated in East Africa and is one of under researched countries. In total, 140 entrepreneurs who own venture of 5-99 employees were able to be interviewed using a survey questionnaire. In this study, structural equation modeling (SEM) was used to examine the direct and indirect relationship of culture in entrepreneurial opportunity exploitation decisions. Findings After hypothesis testing, the empirical results showed that Tanzania’s culture has an impact on entrepreneurial risk taking behavior, which influences entrepreneurial opportunity exploitation decision. It also showed culture through individualism and uncertainty avoidance measurements affect entrepreneurial opportunity exploitation decisions. The empirical results on power distance were insignificant. Research limitations/implications This study is a wake-up call to policy makers and formal institutions such as government authorities, education institutions and religion institutions. Thus, culture has an ability to influence the behavior of entrepreneurs and so the performance of ventures if it is consistent and well structured. Therefore it should be not taken for granted. Data for our study are based on only two cities and therefore the results should not be generalized as the whole country’s inference. Generalizability is questioned because the data are from only two cities in Tanzania and therefore future research should include more cities to be able to validate the generalizability. Practical implications This study is a wake-up call to policymakers and formal institutions such as government authorities, education institutions and religion institutions. Thus culture has an ability to influence the behavior of entrepreneurs and so the performance of ventures if it is consistent and well structured. Therefore it should be not taken for granted. Data for our study are based on only two cities and therefore the results should not be generalized as the whole country’s inference. Social implications In the country which has well-structured culture, influence the behavior of entrepreneurs to exploit opportunities. Originality/value This is the first empirical study to use SEM for exploring the culture of individualism, power distance and uncertainty avoidance impact on entrepreneurial opportunity exploitation in Tanzania.

Predicting customer churn in mobile industry using data mining technology

2017 ◽  
Vol 117 (1) ◽  
pp. 90-109 ◽  
Author(s):  
Eui-Bang Lee ◽  
Jinwha Kim ◽  
Sang-Gun Lee
Keyword(s):  
Data Mining ◽  
Decision Tree ◽  
Online News ◽  
Partial Least Square ◽  
Churn Prediction ◽  
Mining Technology ◽  
Content Type ◽  
Customer Churn ◽  
Mobile Industry ◽  
Using Data

Purpose The purpose of this paper is to identify the influence of the frequency of word exposure on online news based on the availability heuristic concept. So that this is different from most churn prediction studies that focus on subscriber data. Design/methodology/approach This study examined the churn prediction through words presented the previous studies and additionally identified words what churn generate using data mining technology in combination with logistic regression, decision tree graphing, neural network models, and a partial least square (PLS) model. Findings This study found prediction rates similar to those delivered by subscriber data-based analyses. In addition, because previous studies do not clearly suggest the effects of the factors, this study uses decision tree graphing and PLS modeling to identify which words deliver positive or negative influences. Originality/value These findings imply an expansion of churn prediction, advertising effect, and various psychological studies. It also proposes concrete ideas to advance the competitive advantage of companies, which not only helps corporate development, but also improves industry-wide efficiency.

Spam classification: a comparative analysis of different boosted decision tree approaches

2018 ◽  
Vol 20 (3) ◽  
pp. 298-105 ◽  
Author(s):  
Shrawan Kumar Trivedi ◽  
Prabin Kumar Panigrahi
Keyword(s):  
Decision Tree ◽  
False Positive ◽  
False Positive Rate ◽  
False Negative ◽  
The Body ◽  
Content Type ◽  
Performance Accuracy ◽  
Tree Classifier ◽  
Boosted Decision Tree ◽  
Email Spam

PurposeEmail spam classification is now becoming a challenging area in the domain of text classification. Precise and robust classifiers are not only judged by classification accuracy but also by sensitivity (correctly classified legitimate emails) and specificity (correctly classified unsolicited emails) towards the accurate classification, captured by both false positive and false negative rates. This paper aims to present a comparative study between various decision tree classifiers (such as AD tree, decision stump and REP tree) with/without different boosting algorithms (bagging, boosting with re-sample and AdaBoost).Design/methodology/approachArtificial intelligence and text mining approaches have been incorporated in this study. Each decision tree classifier in this study is tested on informative words/features selected from the two publically available data sets (SpamAssassin and LingSpam) using a greedy step-wise feature search method.FindingsOutcomes of this study show that without boosting, the REP tree provides high performance accuracy with the AD tree ranking as the second-best performer. Decision stump is found to be the under-performing classifier of this study. However, with boosting, the combination of REP tree and AdaBoost compares favourably with other classification models. If the metrics false positive rate and performance accuracy are taken together, AD tree and REP tree with AdaBoost were both found to carry out an effective classification task. Greedy stepwise has proven its worth in this study by selecting a subset of valuable features to identify the correct class of emails.Research limitations/implicationsThis research is focussed on the classification of those email spams that are written in the English language only. The proposed models work with content (words/features) of email data that is mostly found in the body of the mail. Image spam has not been included in this study. Other messages such as short message service or multi-media messaging service were not included in this study.Practical implicationsIn this research, a boosted decision tree approach has been proposed and used to classify email spam and ham files; this is found to be a highly effective approach in comparison with other state-of-the-art modes used in other studies. This classifier may be tested for different applications and may provide new insights for developers and researchers.Originality/valueA comparison of decision tree classifiers with/without ensemble has been presented for spam classification.

scholarly journals Motor Programs as Indicators of Penalty Direction in Soccer

2020 ◽  
Vol 8 (9) ◽  
pp. 224-233
Author(s):  
Luiz Felipe Pinto Oliveira da Motta ◽  
Ricardo Fontes Macedo ◽  
Elizabeth Cárpio Rivera ◽  
Angela Luciana De-Bortoli ◽  
Robelius De-Bortoli
Keyword(s):  
Behavior Pattern ◽  
Video Camera ◽  
The Body ◽  
Motor Patterns ◽  
Motor Programs ◽  
Front View ◽  
Elbow Angle ◽  
Patterns Of Behavior ◽  
Prior Intention ◽  
Leg Position

Introduction: Many football games are decided on penalties and usually in championship final games. When seeking to anticipate movements, differences in amplitude can harm players because the informational movement appears to be spread "globally" throughout the action and should be coded at several levels. Thus, it would be interesting to analyze the entire period of the kick, since the player begins his run to approach the ball to recognize the motor patterns used in the kick that identify his direction. Objective: The objective of this study is to identify patterns of behavior in penalty kicks that may indicate the direction of their action / kick and in my moment they appear. Methodology: The sample consisted of 21 subjects hitting a penalty kick, 18 males and 3 females with an average age of 22.18 ± 2.44 years and two goalkeepers with college football experience. The tests consisted of a battery of two penalty kicks for each subject in order to score. The kicks were recorded by a video camera with a front view of the goal goal and the back of the batter. The kick phases were divided into the starting leg position; first step leg; angle of the elbow in relation to the body seen from behind; angle of the elbow in relation to the displacement line seen from above; direction of the tip of the supporting foot and position on the goal where the kick was. Each kick was preceded by a start signal. The data were analyzed from the registration of each variable and the position of the goal in which the ball was kicked, considering it in three sectors: left, right and central. Results: The main results indicated that the variable “Leg of the first step” had 81% of the kicks associated with the direction of the goal; 52.4% of second kicks had repeated the pattern of behavior and 84.6% had repeated the pattern of behavior regardless of the goal position. The variable “Elbow angle in relation to the body seen from behind” had 81.8% repeated behavior pattern regardless of the goal position and the variable “Elbow angle in relation to the displacement line seen from above” had 81% association with the sector of the goal in which the ball was kicked; 52.4% of second kicks had repeated the pattern of behavior and 91.7% had repeated the pattern of behavior regardless of the goal position. Conclusions: The main conclusions indicate that it is possible to relate the kick location with the batter's body information; the start of the race seems to indicate that there is a prior intention of movement programs; the decision of where to hit the penalty appears to be made before contact with the ball and more closely to the placement of the support foot and with this relationship, the size of the goal to be defended by the goalkeeper could be reduced, increasing the possibility of defense.

scholarly journals Banks behavior patterns as a response to the population financial activity in the macroeconomic shocks in Russia

2021 ◽  
Vol 50 (2) ◽  
pp. 74-95
Author(s):  
Yu.S. Evlakhova ◽  
◽  
E.N. Alifanova ◽  
A.A. Tregubova ◽  
◽  
...  
Keyword(s):  
Systemic Risk ◽  
Banking Sector ◽  
Behavior Pattern ◽  
Behavior Patterns ◽  
Economic Shocks ◽  
Macroeconomic Shocks ◽  
Financial Activity ◽  
Risk Research ◽  
Systemically Important Banks

This paper finds out the behavior patterns of the Russian banking sector and systemically important banks in response to changes in the population financial activity under the economic shocks. The results show that the Russian banking sector has a behavior pattern that includes the sequence of actions: the outflow of deposits — vulnerability to non-repayment of loans — deposit bubble — credit bubble. We find no consistent evidence that systemically important banks show the same sequence of actions during the crises. We also find that the banking sector behavior and systemically important banks’ behavior varied in 2008–2009, but became the same in the crisis of 2014–2015. The coincidence of behavior patterns of the banking sector and systemically important banks increases the systemic risk. Research on intragroup differences between systemically important banks will allow finding solutions to reduce the risk.

Optimising pin-in-paste technology using gradient boosted decision trees

2018 ◽  
Vol 30 (3) ◽  
pp. 164-170 ◽  
Author(s):  
Péter Martinek ◽  
Oliver Krammer
Keyword(s):  
Decision Tree ◽  
Prediction Method ◽  
Percentage Error ◽  
Solder Paste ◽  
Hole Filling ◽  
Decision Tree Learning ◽  
Stencil Printing ◽  
Content Type ◽  
Learning Techniques ◽  
Through Hole

Purpose This paper aims to present a robust prediction method for estimating the quality of electronic products assembled with pin-in-paste soldering technology. A specific board quality factor was also defined which describes the expected yield of the board assembly. Design/methodology/approach Experiments were performed to obtain the required input data for developing a prediction method based on decision tree learning techniques. A Type 4 lead-free solder paste (particle size 20–38 µm) was deposited by stencil printing with different printing speeds (from 20 mm/s to 70 mm/s) into the through-holes (0.8 mm, 1 mm, 1.1 mm, 1.4 mm) of an FR4 board. Hole-filling was investigated with X-ray analyses. Three test cases were evaluated. Findings The optimal parameters of the algorithm were determined as: subsample is 0.5, learning rate is 0.001, maximum tree depth is 6 and boosting iteration is 10,000. The mean absolute error, root mean square error and mean absolute percentage error resulted in 0.024, 0.03 and 3.5, respectively, on average for the prediction of the hole-filling value, based on the printing speed and hole-diameter after optimisation. Our method is able to predict the hole-filling in pin-in-paste technology for different through-hole diameters. Originality/value No research works are available in current literature regarding machine learning techniques for pin-in-paste technology. Therefore, we decided to develop a method using decision tree learning techniques for supporting the design of the stencil printing process for through-hole components and pin-in-paste technology. The first pass yield of the assembly can be enhanced, and the reflow soldering failures of pin-in-paste technology can be significantly reduced.

Sign in / Sign up

Export Citation Format

Share Document