Search

Herpes is a type of skin disease characterized by unilateral radicular pain and the appearance of vesicular lesions limited to the skin area. This c4.5 algorithm method is included in the decision tree algorithm to build computational models to predict accurately. According to statistical research ever released by the World Health Organization, one in six people has herpes. Estimates from this study show that about 67% of people worldwide have the herpes virus. Estimates from this study show that about 67% of people worldwide have the herpes virus. Prediction and early detection to reduce the risk of exposure to other people so that medical action can be taken for healing. This prediction is carried out using the Decision tree c4.5 method. This research predicts herpes using eight variables, namely Chlamydia, sugar, alcohol, hepatitis, pregnancy, HIV, Ethylparaben, and Butylparaben. Based on the test results, it was found that the resulting model used the decision tree method, which was 90%. We also recommend using other ways to increase or as a comparison of. [ABSTRACT FROM AUTHOR]

This study utilized Machine Learning (ML) models to analyze and predict customer churns in the context of webbased collaboration platform. Secondary data was used in the study, containing 109,740 observations and 12 predictors. This was divided into test dataset and out-of-time (OOT) dataset where the prior was used for model fitting and the latter is to test performance stability in unseen data. Furthermore, Synthetic Minority Over-sampling Technique (SMOTE) was performed to resolve the data imbalance, hence, preventing bias and distortion in the models' performance. These 3 ML models were assessed based on Accuracy, ROC-AUC, Precision, Recall and F1-Score. Given the business context's applicability, F1-score and Accuracy were used as bases for performance, leading to the selection of Decision Tree Classifier as the ML model in this study, with Accuracy of 92.1% and F1-Score of 63%. Furthermore, hyper-parameter tuning was performed on Decision Tree Classifier to prevent overfitting. To reinforce the model selected, Survival Analysis was implemented, specifically, Kaplan-Meier (KM) Estimator and Cox Proportional Hazard (CPH) were utilized to analyze the rate and timeframe of disengagement to the platform, revealing that beyond 72 months, it was projected to retain only 60% of its user base. Hence, these multidimensional results and insights derived from both Decision Tree Classifier and Survival Analysis were anchored in the formulation of customer retention strategy, proactively target customers who are predicted to churn. [ABSTRACT FROM AUTHOR]

The Digital Markets Act (DMA) constitutes a crucial part of the European legislative framework addressing the dominance of 'Big Tech'. It intends to foster fairness and competition in Europe's digital platform economy by imposing obligations on 'gatekeepers' to share end-user-related information with business users. Yet, this may involve the processing of personal data subject to the General Data Protection Regulation (GDPR). The obligation to provide access to personal data in a GDPR-compliant manner poses a regulatory and technical challenge and can serve as a justification for gatekeepers to refrain from data sharing. In this research-in-progress paper, we analyze key tensions between the DMA and the GDPR through the paradox perspective. We argue through a task-technology fit approach how privacyenhancing technologies - particularly anonymization techniques - and portability could help mediate tensions between data sharing and privacy. Our contribution provides theoretical and practical insights to facilitate legal compliance. [ABSTRACT FROM AUTHOR]

Copyright of CISTI (Iberian Conference on Information Systems & Technologies / Conferência Ibérica de Sistemas e Tecnologias de Informação) Proceedings is the property of Conferencia Iberica de Sistemas Tecnologia de Informacao and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The present research work explores historic data from NCAA men’s basketball datasets with the aim of providing decision-makers with relevant information and improving their judgment when hiring. However, such datasets may contain redundant, noisy, and irrelevant features that could potentially have a negative influence on decision-making processes. In particular, we have addressed a feature selection question to identify player attributes that contribute the most to being chosen by an NBA team for a professional contract. In this context, this paper proposes a data mining approach using a feature selection method to identify relevant characteristics and rank features to assist stakeholders in decision-making processes. Thus, we have used a data mining approach in terms of a feature selection method by testing some models and performing a combination of genetic algorithm with decision tree and SVM. Feature selection is an important step in the data mining process that allows for obtaining better results using a lower number of features. To this end, we performed an experimental study to evaluate the performance of the proposed method with datasets from the NCAA repository, mainly comparing it with conventional feature selection algorithms from the categories Wrapper, Filter, and Embedding, using SVM and Decision Tree as classifiers. The results show that the proposed feature selection technique outperforms the other used methods on this feature selection problem. The results also show that in this specific model, the dataset was reduced from 65 features to only 6, and metrics such as accuracy, recall, precision, and F1-score showed the best values for both the proposed feature selection technique. Furthermore, we found that dbpm, treb, and ogbpm are the three best predictors to determine the tendency of a player to reach the NBA. [ABSTRACT FROM AUTHOR]

The article focuses on using a decision-tree model to predict social media conversion rates, specifically investigating the relationship between advertising investment, customer conversion, and various factors. The research finds that the click-through rate and ad bidding are significant predictors of customer conversion, and while increased spending on social media ads improves conversion rates, there is an exponential increase in conversion rates when advertisers bid above a certain amount.

In 1994, the Nelson Mandela government identified the lack of housing as the most severe social problem faced by South Africa. However, housing information is often inaccessible to those affected by inadequate low-cost housing. The right information at the right time has the potential to not only address the inadequacy of low-cost housing but also to address, in part, other social ills such as unemployment, poverty and government corruption, as well as larger issues such as the climate emergency. This paper presents a conceptual decision tree to govern the knowledge management of architectural information for the possible automation of architecture. Knowledge management, big data and machine learning are the precursors of artificial intelligence, a technology that could further aid in addressing the inadequacy of low-cost housing. A decision tree is the first step. For this paper, the information within frames forms the nodes on the conceptual decision tree. This decision tree is presented graphically and tested hypothetically on a low-cost housing unit. The research findings indicate that there is a noteworthy overlap in architectural information of low-cost housing information presented thereby validating the possible benefit of architectural information knowledge management. [ABSTRACT FROM AUTHOR]

The study analyzes urban energy consumption spatio-temporal patterns using agent-based modeling (ABM). The model uses the Local Indicators of Spatial Association (LISA) and the Geographically Weighted Regression (GWR) to overcome the problem of spatial heterogeneous interference in ABM simulation. In addition, the study also establishes features of preferences of various groups in energy consumption via the Decision Tree (DT). The energy consumption prediction process is executed in six modules via the transformation mechanism of living behavior and preferences of 907 agents in the studied area (Taipei City). The results of the better simulation are verified by Root-Mean-Square Error (RMSE) and Mean Absolute Percentage Error (MAPE). The result in 2030 exhibits rising trend for energy consumption in five areas, with most rising agents concentrated in downtown areas, and declining trend for energy consumption in seven areas. The cumulative distribution function (CDF) of energy consumption prediction shows the number of agents exceeding 600 KWH/household, pointing to increase in regional energy consumption. Simulation results point to trend of further concentration of energy consumption in the future. Given possible shortfall in energy supply in the future, the study suggests to build small-scale power plants or introduce energy-conservation strategies in regions with concentrated energy consumptions, to assure stability of energy demand and supply there. [ABSTRACT FROM AUTHOR]

Data is one of the supporting elements for decision-making in all scientific fields, including in legal dogmatics. As a practical science, legal dogmatics is responsible for responding to legal problems in societies, particularly through court decisions. In the Indonesian legal tradition, which is heavily influenced by the civil law system, judges in charge of deciding cases do not always use references in the form of precedents drawn from similar case decisions. Situations like this often attract criticism as they are likely to cause inconsistencies in court decision-making and will ultimately reduce public confidence in the courts. This paper offers a way to overcome this tendency and aims to model judgment decisions from cases that are related to cases through data mining with the decision tree method. The methods used are Decision Tree ID3, with data derived from decisions on corruption cases decided by the Corruption Court in Indonesia over the past ten years. This data-training model has shown variables that can be used as data-training that really needs consistency in decision making in corruption cases in Indonesia. Considering that the judges' decisions are the result of factual as well as normative reasoning, there are many elements that must be anticipated. The judge's inconsistency in interpreting the facts results in the judge's interpretation of the elements of the crime. This disparity is the source of the weakness of the model that is currently presented in this study. One way to overcome this is to increase the data variance. If the researcher only uses duplicate data from the previous training data, the end result will not be very different. [ABSTRACT FROM AUTHOR]

The C4.5 algorithm is one of many data mining classification techniques. The final result in the classification of this algorithm is the creation of a decision tree. This tree helps in making predictions and selection. The c4.5 algorithm is widely used in many cases, such as outpatient services, disease prediction, etc. It's just that for the c4.5 algorithm method in the construction of chatbots, only a few apply it. More cases of implementing chatbots using Semantic Web, NLP, etc. This study focuses on the c4.5 algorithm method and its application in chatbots because, in previous studies, it was still rare to implement chatbot services in hospitals using this algorithm. The c4.5 algorithm method in this study will be carried out on a chatbot because in predicting poly hospital services for selection, it is more suitable to use this algorithm. The data used were obtained from the Azra Hospital Bogor, totaling 3,292 data after the data cleaning process with the classification results using the c4.5 algorithm, namely getting the RMSE (Root Mean Square) / Mean value of 12.66%. This number means that the lower the average prediction error, the less. That means that from the total data used, 315 data experienced wrong predictions, and 2,977 other data were predicted correctly. Gradient Boosting was also carried out to reduce the prediction error rate by 0.66% after being tested seven times. In the future, this research is expected to improve the classification test time. [ABSTRACT FROM AUTHOR]

Heart disease is the leading cause of death in most countries. According to 2018 statistics, about 17.9 million people worldwide died from cardiovascular disease. Early prediction and control can help reduce mortality from heart disease using existing health data. The decision tree method accurately builds a computational model that aims to predict. The usefulness of decision trees in health has made them used to predict cancer, diagnose lung disease, diagnose heart disease, etc. More studies have also shown that geographical variables can affect the prevalence of hypertension, one of the causes of heart disease. Although hypertension is closely related to geographic variables and has a relationship with heart disease, there are still not many studies related to heart disease prevention and have not used the geographic location as a variable. This study predicts the risk of heart disease by combining 15 variables of heart disease patient data obtained from Azra Hospital, Indonesia and geographic data in the form of an area's altitude. We have categorized the elevation data from 367 villages in West Java into 2 categories based on the indicators of plains in Indonesia, highlands (>600m) and lowlands (<600m). Based on the test results, we found that the decision tree method's diagnosis had an accuracy of 93,75%, with the highest level of risk being in people living in the lowlands. This shows that the altitude of an area can affects the risk of heart disease. We suggest optimizing the method to improve the accuracy of the prediction results. [ABSTRACT FROM AUTHOR]

An effective way to solve the problem of urban traffic congestion is to predict the road traffic status accurately and take effective traffic control measures in time. Considering the impact of visibility on traffic, the pavement status and time characteristics were finely divided, and a regression decision tree was used to establish the traffic flow velocity prediction model with pavement status, time characteristics, and working day characteristics as characteristic parameters. Furthermore, based on the perspective of avoiding using velocity as a single parameter to classify the road traffic status levels, the Kmeans clustering algorithm was used to obtain the classification label results. Moreover, the traffic flow velocity and pavement status were used as characteristic parameters of the classification decision tree to establish the multiparameter road traffic status prediction model. The experimental result showed that the prediction accuracy of the proposed road traffic status prediction model was 81.31%, and this method has good applicability and certain application value for road traffic status prediction. [ABSTRACT FROM AUTHOR]

Machine learning provides various algorithms for application in different domains. In the educational field, huge volume of students' data is generated and machine learning algorithms serve as valuable tool for pattern identification in students' behavior. In this paper, CRISP DM standard for data mining is applied in the research with decision tree algorithm used for modelling on Croatian dataset to develop predictive models for students' outcomes prediction. Data set consisted of 264 students of largest Croatian university collected by online survey. The results prove that decision tree modelling achieves superior results in terms of high accuracy and reliability together with interpretability of tree structure and obtained rules in prediction of students' academic performance. This approach shows promise to be used in student success prediction in the universities in an automatic manner. Such model can be used to: (i) improve students' learning and develop personalized recommender systems for optimal learning paths, (ii) emphasize to professors most important determinants of students' academic success (iii) help management of higher education institutions to facilitate the provision of detailed student learning and adjust institutions strategies, (iv) automate adaptation of the course modules and faculty programs. [ABSTRACT FROM AUTHOR]

This paper identifies the particular difficulties faced by small and medium-sized enterprises (SMEs) in the product development process (PDP) and proposes concrete smart data solutions to address them. The results show a thorough evaluation of general smart data approaches as well as a particular mapping to the problems of PDP. With the help of a decision tree, SMEs in the PDP can use smart data technologies in a targeted and appropriate manner. SMEs must constantly reposition themselves in light of the current challenges posed by megatrends and Industry 4.0 in order to remain competitive. In the future, SMEs will have to innovate faster, adapt more quickly to changes and respond better to customer needs. In particular, Smart Data as a current research topic in the context of Industry 4.0 raises new problems for the design of PDP in the manufacturing industry. [ABSTRACT FROM AUTHOR]

Due to the large amount of sensitive data generated by websites, it is possible to understand the progress of attacks to their databases. This work proposes an intrusion detection system based on data mining and machine learning techniques to detect and mitigate the damage caused by these attacks. It adopts the Information Gain method of selecting attributes in order to reduce the model-building time without affecting the classification performance. Using the CIC-IDS 2017 dataset, this work shows how different decision tree algorithms (Random Forest and J48 Algorithm) behave even if they receive equal parameters and data. Using Information Gain to select attributes, the proposed system achieves a processing time reduction of up to 90%. [ABSTRACT FROM AUTHOR]

Machine learning is a software tool that allows information present in databases to be converted into intelligent decisions based on data and not on experience or feelings. This work presents a review of the algorithms used in the classification of student performance present in the database of the national technical university, specifically in the software engineering career. Information classification analysis using machine learning algorithm has become a modern and powerful instrument, which could help academic institutions improve retention and performance rate of courses by students and with this obtain a more general view of student performance before starting the course based on information from other courses and general student variables in order to reduce the risk of failure. The main objective of this work is to describe the variables used by the National Technical University and the application of different automated learning algorithms in order to obtain the metrics that allow demonstrating the best algorithm among those studied. Finally, the article concludes with a classification algorithm that provides an accuracy of around 80% to 82% that can be replicated in other university institutions. [ABSTRACT FROM AUTHOR]

This paper identifies the particular difficulties faced by small and medium-sized enterprises (SMEs) in the product development process (PDP) and proposes concrete smart data solutions to address them. The results show a thorough evaluation of general smart data approaches as well as a particular mapping to the problems of PDP. With the help of a decision tree, SMEs in the PDP can use smart data technologies in a targeted and appropriate manner. SMEs must constantly reposition themselves in light of the current challenges posed by megatrends and Industry 4.0 in order to remain competitive. In the future, SMEs will have to innovate faster, adapt more quickly to changes and respond better to customer needs. In particular, Smart Data as a current research topic in the context of Industry 4.0 raises new problems for the design of PDP in the manufacturing industry. [ABSTRACT FROM AUTHOR]

Banks have various services offered to their customers, one of which is credit. Credit is a service that provides loans to customers with payment terms and agreements. In the process, these services often experience problems in terms of the repayment process. So, this study aims to look at customer profile data and classify it into two groups, namely customers who have and do not have the opportunity to apply for the loans. This process is carried out with the help of two algorithms, namely the Decision Tree and Artificial Neural Network (ANN), by identifying the customer profile criteria obtained from the prediction of credit offers to customers. Both algorithms are evaluated using three parameters, namely gain ratio, confusion matrix, and Receiver Operating Characteristic (ROC). The Decision Tree produces the highest accuracy performance value of 99,34%, and AUC of 0, 998. Meanwhile, the ANN algorithm produces the highest accuracy of 99,13%, and AUC of 0.969. From each of these algorithms, it can be known about the effect of the treatment carried out on each parameter, and the effect on the performance results obtained. [ABSTRACT FROM AUTHOR]

Many organizations use business processes as a tool to realize and sustain competitive advantage in the market. A business process is a structured collection of activities with comprehensible sequence and dependency to yield a required outcome. The optimization of these processes is of paramount importance because optimized processes yield adaptability, accurate information, enhanced efficiency, accountability through performance monitoring, and improved quality. Relying on business people such as executives and management to identify areas of improvement in the business processes is potentially subjective. This research commences on the assumption that business processes are fully constituted for a business and on this premise seeks an alternate, none subjective, optimization technique. A Decision Tree (DT) is a tool that supports decision making by means of a tree-structured modeling approach to map possible outcomes of a chain of interconnected choices. When applied in statistical regression modeling, a DT model employs supervised learning techniques to model decisions in a tree structure with possible results, input costs, and usefulness. In a DT model, aspects of an element are monitored and the model is trained to predict the future. DT can be applied to improve business processes by identifying activities or elements with significant impact when enhanced. This paper demonstrates business process optimization via DT regression modeling by the use of Python programming. [ABSTRACT FROM AUTHOR]

Before Covid19, airport around the world are faced with limited capacity issues and predicting the passenger demand at the airport is vital for resource planning at various customer touch points ranging from check-in counters and immigration. In this case study, we use one-year historical data to forecast the departure demand for each airline and develop a simulation model to determine the number of check-in counters required to optimize the resource utilization at the airport. The overall objective of this paper is to uncover the real-world business problem in the classroom setting and the authors will share how to go through this case study in detail. [ABSTRACT FROM AUTHOR]

Instrumental helping has been reported in infants toward other humans but not toward robots. Providing infants with opportunities for action-based assistance to robots might lead to more efficient infant-robot interactions. This paper presents preliminary findings on infants' spontaneous instrumental helping to robots exhibiting motion challenges, and proposes a novel decision-making model for infant-robot interaction that encompasses instrumental helping in its parameters; both in the context of pediatric rehabilitation. Six infants were engaged in a chasing game with a wheeled robot with the goal to follow the robot and ascend an inclined platform (8 sessions, 4 weeks). After infants' instrumental helping toward the robot was identified, a decision tree model was created to evaluate a set of annotated variables as potential predictors to the observed behavior. Next, a Markovian model for robot control was developed where these predictors were used as parameters to promote, in turn, action-based goals for the infants. [ABSTRACT FROM AUTHOR]

Naïve Bayes Classifier (NBC) is one of the most popular machine learning methods that has been applied in various fields, some of them being text classification, medical diagnosis and systems performance management. Claims about the excellence of this classifier’s performance has been stated in several literatures. The main goal of this study is to assess the consistency of this classifier’s performance by applying it to five medical datasets and compare the result with another popular classification method, Decision Tree (DT). Results present some empirical proofs that NBC has a consistent performance, performs better compared to DT by outperforming it in four out of the five datasets that are used and is robust to the presence of missing value. This consistent performance may be due to the nature and background of the datasets used, in which datasets with the right attribute variables tend to produce a high true positive rate and accuracy score. [ABSTRACT FROM AUTHOR]

Parkinson’s Disease has two subtypes which are Tremor Dominant (TD) and Postural Instability/Gait Difficulty (PIGD). Each subtype has the difference in clinical treatment, so it is necessary to classify Parkinson’s Disease subtypes. Three classification methods were implemented: decision tree, logistic regression, and logit leaf model (LLM). Data on 229 people with early Parkinson’s disease from the PPMI (Parkinson’s Progression Markers Initiative) database were used in the analysis. Imbalanced data problem were handled using oversampling, undersampling, SMOTE (Synthetic Minority Over-sampling Technique). Logistic regression with SMOTE using parameter set-up α=600, γ=200 produced the best result, according to the accuracy of 98.3 %, sensitivity of 98.41 %, and specificity of 99.07 %. [ABSTRACT FROM AUTHOR]

Every year more than a million women are diagnosed with breast cancer that results in the death of more than half of them, due to the delay in diagnosis of the disease. Machine Learning (ML) offers a better approach for the prediction of breast cancer. A feature selection technique, INTERACT is applied to select relevant features for breast cancer diagnosis. The tool used in this project is WEKA 2.3. The classification algorithms, namely Decision tree, Random forest and Support Vector Machine (SVM) are incorporated along with INTERACT technique. The relevant and important features for breast cancer diagnosis are chosen. The dataset used in this project is Wisconsin Diagnostic Breast Cancer (WDBC) subdirectory. The experimental result shows that the SVM outperforms the other classifiers. It improved accuracy of the diagnostic model by using feature selection method. [ABSTRACT FROM AUTHOR]

Copyright of KnE Engineering is the property of Knowledge E DMCC and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Gamification has been used in a wide variety of subject-specific education contexts. Examples of such usage in the Supply Chain Management (SCM) context include the oft-played beer distribution game, developed by MIT Sloan School of Management (Forrester, 1961), which simulates the coordination of typical problems in supply chain processes, promoting information sharing and collaboration throughout a supply chain (Sterman, 1984). Purchasing and Supply Management (PSM), a subset of this wider SCM area, focuses on the direct relationships between organisational buyers and suppliers, covering aspects such as establishing trust, identifying and selecting suitable suppliers, managing supplier performance and the overall relationship. A systematic review of the PSM gamified learning literature establishes that there has been limited research to date and that which there is tends to focus on quantitative representations of managing overall supply and demand, using wider SCM elements. This suggests that there are opportunities to gamify PSM learning, in particular focusing on the human element in PSM and developing soft skills, as strong buyer-supplier relationships can generate significant benefits to both parties. To provide a more focused PSM contribution, a second systematic literature review distils the relevant principles, techniques and processes to inform the development of two gamified PSM learning activities. Negotiation and supplier relationship management rely heavily on personal interactions and are both seen as key activities at different stages of the PSM process. The development of the two gamified learning activities is strengthened by being underpinned by a synthesis of the literature review’s key findings, ensuring they are domain-meaningful abstractions of reality, contain rewards and rankings based on clear objectives and have appealing gameplay. It is hoped that this paper provides a platform for future domain specific PSM research and will be of use to educators in this field in developing their own gamified learning. [ABSTRACT FROM AUTHOR]

The “Predicting a Pulsar Star” dataset has been investigated in this paper. The pulsars are special neutron stars emitting narrow beams of light with high energy into space. They rotate very rapidly and the signal repeats. Scientists seek periodic radio signals in order to detect these objects. Of course, frequency patterns vary from one star to another. The signal is averaged on a lot of rotations. At the same time a lot of pulsar candidates are just radio noise. There is a lot of objects to check. In this task machine learning algorithms can be applied to remove objects that definitely do not belong to pulsars. This step allows reducing time to handle signals of potential pulsars. Astronomers could check only “difficult” cases for classification. Other objects are supposed as pulsars registered with high probability. Review of literature has shown that usually only one algorithm is used in each paper on astronomy objects detection or too complex approaches are applied. For example, training neural nets takes a lot of time and it needs attention of specialists. In this research simple classification algorithms are applied to this task: the Naive Bayes, the logistic regression and the CART decision tree classifiers. The ensemble methods are applied at the extended dataset and its transformed version by means of principal component analysis with polynomial kernels. The random forest ensemble method is used. Averaged accuracy values of the constructed classifiers are about 93%. Better accuracy cannot be achieved because of high level of noise in the dataset. [ABSTRACT FROM AUTHOR]

Inspection lines are a heavy duty yet expensive part in many industries such as wireless communication industry, auto industry. While inspection is always a necessary evil in a very complicated and competitive industry, there is a strong need to reduce the unnecessary inspection processes and resources to improve the competitiveness. This work develops a framework to apply big data analytics to improve the inspection processes based on case study method. A world class case company is selected and the data-warehouse is established for a specific lot of products. After the data pre-processing work, the relationships among the detection stations and the detection results could be found. Various meta-models are used, which including association rule method to find the correlation between different test items and to reduce the number of required tests. Decision tree is also applied to find the optimal controllable factor. Finally, the developed rules are further verified by empirical data. Based on the empirical results, the number of detection items could be reduced by around 16.2% - a huge cost saving. It is also found out that the critical control factors, recommended by the decision tree, are temperature and humidity, which is a way to improve its quality without extra cost. The proposed big data application framework is therefore feasible in this case and machine learning or other models could be further extended, which is the future research direction. [ABSTRACT FROM AUTHOR]

Excessive drinking by workers is linked to the manifestation of noncommunicable chronic diseases, such as metabolic syndrome. The role of occupational health nurses is to encourage workers to adopt health behaviors to avoid excessive drinking. The study aimed to conduct an exploratory examination of the factors related to Japanese workers’ maintenance of the health behavior of avoiding excessive drinking. The study was conducted using a longitudinal questionnaire on the Internet. The survey was conducted once in 2014 and once in 2015. The questionnaire items included whether the respondent was engaged in health behaviors, such as avoiding excessive drinking, the degree of the influence of stress on the diet, the Daily Hassles Scale, and the Resilience Scale. The responses were analyzed using a decision tree analysis. The dependent variable was whether or not the respondent continued to avoid excessive drinking in both the 2014 and 2015 surveys. The independent variables were the questions in 2014, and the relationships between the dependent variable and independent variables were examined. The total number of respondents was 359, and the average age was 51.4 ± 7.6 years. Of these, 85.8 percent were male, 94.4 percent were in full-time employment, 90.8 percent worked during the day, and 61.0 percent were office workers. Those who continued to avoid excessive drinking constituted 56.0 percent. The decision tree analysis found that the continuation of excessive drinking avoidance was most strongly correlated with the degree of heavy drinking due to stress and it correlated with other factors, including the degree of overeating due to stress, trying to eat more vegetables, not smoking, and not snacking (error estimate 0.27; cross-validation estimate 0.33). The continuation of the excessive drinking avoidance behavior was most strongly correlated with the degree of heavy drinking due to stress. Occupational health nurses need to introduce appropriate knowledge about self-care when providing health instructions regarding lifestyle issues. [ABSTRACT FROM AUTHOR]

Migration movement is one of the main drivers of modern economic development. Due to the influence of human factor on the economic system development, the attempts of describing, estimation and modeling this phenomenon are characterized by relatively low accuracy. According to one of the current trends of individual economic behavior description it is necessary to describe the decision making of a person from the position of a rational individual. The second trend consists in estimating individual behavior from the position of irrational decision making and describe the variants of system development using an approach, which takes into account as the main factor the previous system experience. One of the goals of our research was to estimate migrants behavior in the hosting country and to analyze their decisions on continuing their work as a migrant employee. In this paper we use an approach for migration description, that is based on machine learning methods, in particular the decision tree and the k-nearest neighbor methods. The best accuracy of estimation during our work was observed when using the decision tree method. According to our approach 8 percent of migrants wish to stay for work in Russia for additional three months, 35 percent for an additional year and 57 percent wish to stay for a time period from one to three years, which can be explained by a high attractiveness of well developed labor markets for migrants. [ABSTRACT FROM AUTHOR]

Due to the large user base in social media, social media posts and comments hold discussions on trending issues. Different Government and Non-Government organizations actively maintain social media accounts where many people engage and discuss their problems. This provides a great opportunity to systematically arrange and present these problems topic-wise which helps to direct these into appropriate recipients. Previous works provide techniques on topic extraction, sentiment classification, opinion mining from natural language text. However, these techniques are neither applied to social media nor do consider the multilingual aspect of social media. In this paper, a learning-based technique is proposed that extracts important keywords from multilingual social media interactions and provides a ranked list of these under predefined topics. The proposed technique is evaluated on Bangla, English and French comments from well-known and/or verified Facebook pages. It was observed that a high accuracy both in case of keyword extraction and ranking can be achieved by utilizing the proposed approach. [ABSTRACT FROM AUTHOR]

Copyright of CISTI (Iberian Conference on Information Systems & Technologies / Conferência Ibérica de Sistemas e Tecnologias de Informação) Proceedings is the property of Conferencia Iberica de Sistemas Tecnologia de Informacao and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

This study intends to estimate the profit of a movie through the construction of a predictive model that uses several Data Mining techniques, namely neural networks, regression and decision trees. The model will allow obtaining the prediction of box office revenue. Three different dependent variable approaches were used (interval, categorical and binary) aiming to study the difference and predictive influence that each one has on the results. Two metrics were used to determine the most accurate predictions: the misclassification error for the categorical models and the average squared error for the continuous one. In this study, the best predictive results were obtained through the use of multi-layer perceptron. Regarding the representative distinction between the dependent variable, the multiclass model presents a much higher error rate comparing to the remaining ones, which is explained with the increase of the number of classes to predict. [ABSTRACT FROM AUTHOR]

Approximate computing has recently reemerged as a design solution for additional performance and energy improvements at the cost of output quality. In this paper, we propose using a tree-based classification algorithm as an approximation tool for general-purpose applications. We show that, without any hardware support, completely implemented in software, our approach can improve performance by up to 4x (1.95x on average) and reduce EDP by up to 19x (4.04 on average) when compared to precise executions. Besides that, in some cases, our software-based mechanism can even outperform traditional hardware-based Neural Network's state-of-the-art designs. [ABSTRACT FROM AUTHOR]

In the work we consider the problem of optimal organization of healthcare system taking into account both map and statistical data. Map data include degrees of urbanization which classify the character of an area. This approach is based on mapping the territory by a grid square cell of 1 km2. Data about public health include research concerning spreading of basic infectious diseases and distribution of hospital system. Decision making is based on application of decision tree induction algorithm for tuples containing both numerical and categorized data. Map data were aggregated before they were used in learning tuples. Statistical data concerning public health were categorized with help of quantile classification method. We use a collection of functions of R software environment to visualize spatial urbanization data and models on top of static maps from various online sources. It includes tools common to those tasks, including functions for aggregation of spatial data and implementation of algorithm of decision tree induction in C5.0 package. [ABSTRACT FROM AUTHOR]

The relationship between ownership structure (private vs State-owned) and company performance has been deeply analyzed by scholars and practitioners. Prior studies found mixed results about this topic; some scholars demonstrated that private firms perform better than State-Owned Enterprises and others came to opposite or undefined results. Further, during the global financial crisis, this topic gained relevance. To the best of our knowledge, Italian framework suffers of a lack of these studies. Furthermore, no studies classify private and State-owned firms on the basis of the decision tree approach. The main aim of this paper is thus to classify private Italian listed companies and State-owned Italian listed companies by using a decision tree classification approach. To do so, we performed a decision tree classification analysis on a sample of 18 State-owned listed firms and 212 private listed firms. Empirical results show that the most relevant financial values that stakeholders can rely in order to classify private and State-owned Italian listed firms are total debt, net sales and total assets. [ABSTRACT FROM AUTHOR]

Researching entrepreneurial intention has become extremely popular due to the importance entrepreneurial activity has for the economy. Data mining has been used increasingly in the realms of prediction and classification, replacing traditional statistical approaches. The aim of this paper is to analyse the usefulness of decision trees for predicting entrepreneurial intention by comparing the decision tree method with structural equation modelling (SEM). A predictive model was proposed and empirically tested on a sample of 218 business students from Croatia and Macedonia. The results show that both techniques, SEM and decision tree, offer roughly equal theoretical contributions. However, the rules established using decision trees have deepened our understanding of the data by pointing to specific groups of students in the sample and their respective probabilities of having entrepreneurial intentions. [ABSTRACT FROM AUTHOR]

The textile industry is one of the major industries and has a high level of integration for the country's economy. The use of the right method can increase good profits and make it easier for the company. The 4 countries with the largest textile industries include Bangladesh, China, India and Indonesia. Decision Tree is included in data mining that is used by many industries, one of which is the textile industry. The decision tree is in the form of classifying data in the form of classes and nodes, a method that is easy to understand and very capable of providing the effectiveness of a company's production process. This paper uses the C4.5, SVM and Multi-Layer Perceptron algorithms. The methodology used is Decisions Tree which has the right level of accuracy and classification for the textile industry, because it is very efficient, has a variety of models and a structured rule tree structure. This method is used to be implemented in the textile industry, including being able to predict fabric defects, and clustering information in the textile industry. The company wants quality fabric products. Classification is in the form of class and root so that it can determine the right level of accuracy and can determine the level of defects in the fabric whether later the fabric can be processed into finished products or not and inform the level of smoothness. A high level of accuracy is obtained from Percepetron Multi-Layer with a value of 83.6%. [ABSTRACT FROM AUTHOR]

Copyright of Scientific Journal Systemy Wspomagania w Inzynierii Produkcji is the property of P.A. Nova S.A. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of CISTI (Iberian Conference on Information Systems & Technologies / Conferência Ibérica de Sistemas e Tecnologias de Informação) Proceedings is the property of Conferencia Iberica de Sistemas Tecnologia de Informacao and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The report examines the peculiarities of making investment decisions aimed at implementing projects to reduce the risks of natural disasters. This type of decisions is perceived as a specific type of managerial decision. The strategic framework in which decisions are taken is studied. The documents on which these strategic decisions can be based are strategically systematized. The peculiarities of the stakeholders concerned by the decisions and their possible attitude to risk are analyzed. The specificity of the evaluation of possible damages in the event of natural disasters has been identified. The possible benefits of implementing projects to reduce the risks of natural disasters are outlined. The prospects for choosing an investment alternative using the “Decision Tree” and “Influence Diagrams” are analyzed. [ABSTRACT FROM AUTHOR]

Young companies need support concerning decisions related to intellectual properties. Entrepreneurs can resort to a menu of strategies, not only patenting. First, we explore the literature on standardization and patenting and relate it to entrepreneurship to identify the internal and external influencing factors as well as the motives and risks related to decision making. Then, we conduct five case studies to explore these influencing factors, while trying to reconstruct the decision making process. We find five main factors: technology, resources, knowledge protection versus knowledge diffusion, need for partnerships, and innovation dynamics. Companies should use patents when their technology is patentable and knowledge protection is essential for survival. Standardization is suitable when knowledge diffusion is more important than protection, and companies look for establishing new partnerships. These insights are integrated into a decision tree that provides guidance to young entrepreneurs to make an informed decision regarding intellectual properties. [ABSTRACT FROM AUTHOR]

Information theoretic measures with their unique feature of smooth implementation are successfully applied in various areas such as physics, probability and statistics, communication theory and economics etc. They plays important role in data mining by improving the classifying algorithms. Further, they have been successfully used for classification and defining rules of any type of data with ID3 algorithm. In this paper, we have modified ID3 algorithm and used generalized entropy of order α and type β to derive some more refined rules from the resultant decision tree in context with the analysis of the health conditions in the States and Union Territories of India. [ABSTRACT FROM AUTHOR]

In tourist and business transportation by trains or buses at the final destination point it is often necessary for passengers to use a car to perform their daily logistical tasks. The alternative mode of transport are motorail trains, which offer convenience for passengers to travel from door to door in their own car. Accordingly, it is very important to study the effectiveness of the organisation of motorail trains for Bulgarian railway routes along which business and tourist trips are made. Providing such services is a factor in tourist travel and transport development. The experimental research has been carried out for the Sofia-Plovdiv-Burgas railway axis. In this paper a methodology for assessing the effectiveness of the organisation of motorail trains is proposed. It comprises a few steps. In the first step, by using the AHP method, the variants of transportation by motorail trains are formed and the probabilities for stochastic variation of passengers are determined. In the second step the operating costs of transportation and financial investment for a new type of wagons are determined. The payback period has been chosen as a criterion of efficiency. In the third step a Decision Tree Analysis is performed to determine the best variant of transportation. The environmental and social benefits of the new transport technology have been determined. The results, which have been obtained, indicate that the motorail trains would be an effective technology for thatparticular railway route. [ABSTRACT FROM AUTHOR]

In this study, the effect of dimension for a feature vector on the classification of Turkish e-mails as spam or legitimate is investigated. Although hundreds of experimental studies are achieved especially for English, which is a nonagglutinative language, the number of efforts for Turkish, which is one of the most popular agglutinative languages in the world, is counted something on the fingers of one hand. Therefore, a solution is sought for Turkish spam e-mail problem taking the special characteristics of Turkish e-mails into consideration. The developed spam filtering framework has four components named as morphological decomposition, feature selection, training, and test phases. A fixed-prefix stemming approach is used to extract the features of an e-mail and then the Mutual Information (MI) method is carried out as the feature selection method. The Decision Tree (DT) and Artificial Neural Network (ANN) classifiers are employed and the recognition accuracies obtained from these methods are considerably satisfactory. The highest accuracy rates are 91.08% for ANN and 87.67% for DT methods when the dimensions of feature vectors are selected as 150x5) and (75x5), respectively. [ABSTRACT FROM AUTHOR]

This research promises a two class classification of digitized mammogram images using two feature extraction algorithms; Dense Scale Invariant Feature Transform (DSIFT) and Linear Configuration Pattern (LCP). The dataset that is used for this research is retrieved from Image Retrieval in Medical Applications (IRMA) project which provides mammographic patches consisting of normal, benign and malignant cases of publicly available digital mammographic databases. In this research, 200 normal and 200 abnormal lesion cases of Digital Database for Screening Mammography (DDSM) are used. Both linear configuration pattern (LCP) and DSIFT feature vectors have a high accuracy in classifying normal cases versus abnormal up to 100%. This approach can be used in computer-aided diagnosis to help radiologists to distinguish a healthy case from others in an efficient way. [ABSTRACT FROM AUTHOR]

A Decision Tree algorithm (DTA) from data is created using construction criteria and also considering the don't care attributes, for each level of the tree. The DTA goal is to create on-demand a short and accurate decision tree from either data or a stable (or dynamically changing) set of rules. A set of steps provides a decision tree with a definite number of nodes and leaves. The pruning of decision rules case is also examined with the consequences on the accuracy. An improved version of DTA (IDTA) provides smaller DT and eliminates branches by using the criterion of less length (CLL). [ABSTRACT FROM AUTHOR]

Random Forest is an Ensemble Supervised Machine Learning technique. Research work in the area of Random Forest aims at either improving accuracy or improving performance. In this paper we are presenting our research towards improvement in learning time of Random Forest by proposing a new approach called Disjoint Partitioning. In this approach, we are using disjoint partitions of training dataset to train individual base decision trees. This helps in creating diversity in base decision trees. Also different subsets of attributes are used at each node of decision tree to increase diversity. This approach generates Random Forest classifier which is trained efficiently and gives classification accuracy comparable to the original Random Forest approach. [ABSTRACT FROM AUTHOR]