Search results

Classification techniques have been applied to many applications in various fields of sciences. There are several ways of evaluating classification algorithms. The analysis of such metrics and its significance must be interpreted correctly for evaluating different learning algorithms. Most of these measures are scalar metrics and some of them are graphical methods. This paper introduces a detailed overview of the classification assessment measures with the aim of providing the basics of these measures and to show how it works to serve as a comprehensive source for researchers who are interested in this field. This overview starts by highlighting the definition of the confusion matrix in binary and multi-class classification problems. Many classification measures are also explained in details, and the influence of balanced and imbalanced data on each metric is presented. An illustrative example is introduced to show (1) how to calculate these measures in binary and multi-class classification problems, and (2) the robustness of some measures against balanced and imbalanced data. Moreover, some graphical measures such as Receiver operating characteristics (ROC), Precision-Recall, and Detection error trade-off (DET) curves are presented with details. Additionally, in a step-by-step approach, different numerical examples are demonstrated to explain the preprocessing steps of plotting ROC, PR, and DET curves.

In order to estimate the value of semi-automated subject indexing in operative library catalogues, the study aimed to investigate five different automated implementations of an open source software package on a large set of Swedish union catalogue metadata records, with Dewey Decimal Classification (DDC) as the target classification system. It also aimed to contribute to the body of research on aboutness and related challenges in automated subject indexing and evaluation.

On a sample of over 230,000 records with close to 12,000 distinct DDC classes, an open source tool Annif, developed by the National Library of Finland, was applied in the following implementations: lexical algorithm, support vector classifier, fastText, Omikuji Bonsai and an ensemble approach combing the former four. A qualitative study involving two senior catalogue librarians and three students of library and information studies was also conducted to investigate the value and inter-rater agreement of automatically assigned classes, on a sample of 60 records.

The best results were achieved using the ensemble approach that achieved 66.82% accuracy on the three-digit DDC classification task. The qualitative study confirmed earlier studies reporting low inter-rater agreement but also pointed to the potential value of automatically assigned classes as additional access points in information retrieval.

The paper presents an extensive study of automated classification in an operative library catalogue, accompanied by a qualitative study of automated classes. It demonstrates the value of applying semi-automated indexing in operative information retrieval systems.

Previous research has focused on the outcomes of telework, investigating the advantages and disadvantages of teleworking for employees. However, these investigations do not examine whether there are differences between teleworkers when evaluating the advantages and disadvantages of teleworking. The aim of this study is to identify of distinct classes of teleworkers based on the advantages and disadvantages that teleworking has for them.

This study used secondary survey data collected by the Spanish National Statistics Institute (INE). A sample of 842 people was used for this study. To identify the distinct classes of teleworkers, their perceived advantages and disadvantages of teleworking were analyzed using latent class analysis.

Three different classes of teleworkers were distinguished. Furthermore, sociodemographic covariates were incorporated into the latent class model, revealing that the composition of the classes varied in terms of education level, household income, and the amount of time spent on teleworking per week. This study also examined the influence of these emergent classes on employees’ experience of teleworking.

This study contributes to previous research investigating if telework is advantageous or disadvantageous for teleworkers, acknowledging that teleworkers are not identical and may respond differently to teleworking.

This paper focuses on patterns of film consumption within cultural consumption more broadly to assess trends in consumerism such as eclectic consumption, individualised consumption and omnivorous/univorous consumption and whether economic background and status feature in shaping cultural consumption. We focus on film because it is widely consumed, online and offline, and has many genres that vary in terms of perceived artistic and entertainment value. In broad terms, film is differentiated between mainstream commercially driven film such as Hollywood blockbusters, middlebrow “feel good” movies and independent arthouse and foreign language film. Our empirical statistical analysis shows that film consumers watch a wide range of genres. However, films deemed to hold artistic value such as arthouse and foreign language feature as part of broad and wide-ranging pattern of consumption of film that attracts its own dedicated consumers. Though we found that social and economic factors remain predictors of cultural consumption the overall picture is more complex than a simple direct correspondence and perceptions of other cultural forms also play a role. Those likely to consume arthouse and foreign language film consume other film genres and other cultural forms genres and those who “prefer” arthouse and foreign language film have slightly more constrained socio-economic characteristics. Overall, we find that economic and cultural factors such income, education, and wider consumption of culture are significant in patterns of film consumption.

Mobile phone-based human activity recognition (HAR) consists of inferring user’s activity type from the analysis of the inertial mobile sensor data. This paper aims to mainly introduce a new classification approach called adaptive k-nearest neighbors (AKNN) for intelligent HAR using smartphone inertial sensors with a potential real-time implementation on smartphone platform.

The proposed method puts forward several modification on AKNN baseline by using kernel discriminant analysis for feature reduction and hybridizing weighted support vector machines and KNN to tackle imbalanced class data set.

Extensive experiments on a five large scale daily activity recognition data set have been performed to demonstrate the effectiveness of the method in terms of error rate, recall, precision, F1-score and computational/memory resources, with several comparison with state-of-the art methods and other hybridization modes. The results showed that the proposed method can achieve more than 50% improvement in error rate metric and up to 5.6% in F1-score. The training phase is also shown to be reduced by a factor of six compared to baseline, which provides solid assets for smartphone implementation.

This work builds a bridge to already growing work in machine learning related to learning with small data set. Besides, the availability of systems that are able to perform on flight activity recognition on smartphone will have a significant impact in the field of pervasive health care, supporting a variety of practical applications such as elderly care, ambient assisted living and remote monitoring.

The purpose of this study is to build and test an accurate offline model by using only a compact training data that can reduce the computational and memory complexity of the system. This provides grounds for developing new innovative hybridization modes in the context of daily activity recognition and smartphone-based implementation. This study demonstrates that the new AKNN is able to classify the data without any training step because it does not use any model for fitting and only uses memory resources to store the corresponding support vectors.

This study aims to analyse the impact of digital divide (DD) on income inequality in sub-Saharan Africa over the period 2004–2016.

In applying a finite mixture model (FMM) to a sample of 35 sub-Saharan African (SSA) countries, this study posits that DD affects income inequality differently.

The findings show that the effect of DD on income inequality varies across two distinct groups of countries, which differ according to their level of globalization. In addition, the study shows that most globalized countries are more inclined to be in the group where the effect of DD on income inequality is negative. The results are consistent with several robustness checks, including alternative measures of income inequality and additional control variables.

This study complements that extant literature by assessing linkages among the DD, globalization and income inequality in sub-Saharan African countries contingent on cross-country heterogeneity.

This paper aims to propose a solution for detecting and grading diabetic retinopathy (DR) in retinal images using a convolutional neural network (CNN)-based approach. It could classify input retinal images into a normal class or an abnormal class, which would be further split into four stages of abnormalities automatically.

The proposed solution is developed based on a newly proposed CNN architecture, namely, DeepRoot. It consists of one main branch, which is connected by two side branches. The main branch is responsible for the primary feature extractor of both high-level and low-level features of retinal images. Then, the side branches further extract more complex and detailed features from the features outputted from the main branch. They are designed to capture details of small traces of DR in retinal images, using modified zoom-in/zoom-out and attention layers.

The proposed method is trained, validated and tested on the Kaggle dataset. The regularization of the trained model is evaluated using unseen data samples, which were self-collected from a real scenario from a hospital. It achieves a promising performance with a sensitivity of 98.18% under the two classes scenario.

The new CNN-based architecture (i.e. DeepRoot) is introduced with the concept of a multi-branch network. It could assist in solving a problem of an unbalanced dataset, especially when there are common characteristics across different classes (i.e. four stages of DR). Different classes could be outputted at different depths of the network.

Legal education, like any other discipline in higher education, necessitates in use of various teaching and learning pedagogies in order to provide a sustainable teaching and learning experience. This article aims to examine the feasibility of implementing flipped learning method as a pedagogy on legal students at the Open University of Sri Lanka, as well as the perceptions of students and lecturer on the teaching and learning process in a flipped class in preparation for future implementation.

A mixed research method was used. A survey and a semi-structured interview were used to collect student perceptions, and observations of the lecturer were used to document the lecturer's perception.

According to the information gathered from both qualitative and quantitative data, the flipped learning pedagogy enhances the prior learning and student-centered learning of open and distance learning (ODL) and offers a new perspective on the existing pedagogies used in legal education. This article also emphasizes that an equitable implementation of designing and delivering a flipped class will ensure the effectiveness in teaching and learning law in Sri Lanka through ODL.

Despite the fact that there is substantial academic literature on flipped pedagogy, including in legal education, this article will create an original contribution by incorporating reflections from Sri Lankan legal education as well as ODL.

The purpose of this paper is to provide insights for flower retailers, horticultural practitioners and marketing managers into the prioritisation of cut flower attributes by German residents.

Applying a best–worst scaling approach, this analysis identified the relative ranking of importance amongst product attributes relevant to German consumers when buying fresh cut flowers. A latent class analysis determined four flower consumer segments for further study. The study builds on a sample of 978 consumers and is consistent with the most recent German census in terms of age, gender, income and federal state.

The best-worst analysis showed that intrinsic flower attributes, in particular appearance, freshness and scent were found to be more important to German consumers than the extrinsic attributes studied, namely, price, country of origin and a certification indicating fair trade. The latent class analysis determined four consumer segments that desire either budget, luxury or ethical flowers or more information about flowers. For all identified consumer segments, appearance was the attribute of greatest importance. The segments that desired luxury or ethical flowers, as well as the segment that desires more information were interested in appearance, but also had relatively large shares of preferences dedicated to flower freshness guarantees. The preference for freshness guarantees in addition to appearance may be interpreted jointly as a desire for not only beautiful and aesthetically pleasing flowers, but for sustained beauty.

Internationally, the study fills a research gap by exploring consumer’s relative preference for cut flower attributes. In contrast to existing studies on consumer preferences for flowers in Germany, the present study builds on a sample that was targeted in terms of age, gender, net household income and federal state to the most recent German census.

This study aims to investigate the associations among decent work (DW), capabilities and the flourishing of employees in a South African context.

A cross-sectional survey was conducted with a convenience sample (N = 436) of early childhood development practitioners from two South African provinces. A demographic questionnaire, the Decent Work Scale, the Capability Set for Work Questionnaire and the Flourishing-at-Work Scale were administered.

Latent class analysis showed four capability sets: robust, relational, knowledge/skills and weak capability sets. Employees with a robust capability set were more inclined to report DW than those with knowledge/skills and weak capability sets. Employees with a weak capability set were significantly less inclined to report organisational values that complement family and social values than the other three capability sets. Employees with a robust capability set reported significantly higher emotional well-being (EWB), psychological well-being (PWB) and social well-being (SWB) levels than those with relational, knowledge/skills and weak capability sets. DW was significantly related to EWB, PWB and SWB.

This study contributes to the literature regarding DW, capabilities and flourishing of employees in a non-western, educated, industrialized, rich and democratic and non-POSH context. The study highlights the need for well-being policies that focus on DW and the capabilities of people in disadvantaged positions. These together would strengthen their agency for converting capabilities into well-being.