Search results

Although risk culture is a key determinant for an effective risk management, identifying the risk culture of a firm can be challenging due to the abstract concept of culture. This paper proposes a novel approach that uses unsupervised machine learning techniques to identify significant features needed to assess and differentiate between different forms of risk culture.

To convert the unstructured text in our sample of banks' 10K reports into structured data, a two-dimensional dictionary for text mining is built to capture risk culture characteristics and the bank's attitude towards the risk culture characteristics. A principal component analysis (PCA) reduction technique is applied to extract the significant features that define risk culture, before using a K-means unsupervised learning to cluster the reports into distinct risk culture groups.

The PCA identifies uncertainty, litigious and constraining sentiments among risk culture features to be significant in defining the risk culture of banks. Cluster analysis on the PCA factors proposes three distinct risk culture clusters: good, fair and poor. Consistent with regulatory expectations, a good or fair risk culture in banks is characterized by high profitability ratios, bank stability, lower default risk and good governance.

The relationship between culture and risk management can be difficult to study given that it is hard to measure culture from traditional data sources that are messy and diverse. This study offers a better understanding of risk culture using an unsupervised machine learning approach.

Advanced big data analysis and machine learning methods are concurrently used to unleash the value of the data generated by government hotline and help devise intelligent applications including automated process management, standard construction and more accurate dispatched orders to build high-quality government service platforms as more widely data-driven methods are in the process.

In this study, based on the influence of the record specifications of texts related to work orders generated by the government hotline, machine learning tools are implemented and compared to optimize classify dispatching tasks by performing exploratory studies on the hotline work order text, including linguistics analysis of text feature processing, new word discovery, text clustering and text classification.

The complexity of the content of the work order is reduced by applying more standardized writing specifications based on combining text grammar numerical features. So, order dispatch success prediction accuracy rate reaches 89.6 per cent after running the LSTM model.

The proposed method can help improve the current dispatching processes run by the government hotline, better guide staff to standardize the writing format of work orders, improve the accuracy of order dispatching and provide innovative support to the current mechanism.

This study explores digital transformation in the tea supply chain within developing economies, with a focus on smallholder tea producers in Sri Lanka. Tea is one of the most widely consumed beverages in the world. Among the tea producers, smallholder tea producers account for a substantial portion of total tea production in several countries. Mobile phones play a significant role in providing smallholder producers with access to crucial agricultural information, markets and financial services.

This study adopts a deductive approach, analysing mobile phone ownership, literacy, experience and perception among smallholder tea producers. The chi-squared test of independence and hierarchical clustering methods were used to test the hypotheses and address the research questions.

The study identifies four clusters of smallholder tea producers as Basic Tech Adopters, Digital Laggards, Skeptical Feature Phone Users and Tech-savvy Adopters based on their characteristics towards mobile-based technologies. Approximately 75% of the surveyed sample, which included both tech-savvy and basic-tech adopters, showed a positive attitude toward adopting mobile-based agricultural technologies.

The study suggests developing targeted strategies and policies to enhance the productivity of the smallholder tea production process in developing economies. The study highlights the importance of awareness, access, affordability and availability when implementing digital services for businesses at the base of the pyramid, such as tea smallholdings in developing economies.

The present study aims to address the lack of data-driven empirical studies on the use of mobile phones in smallholder settings. The findings of this study enable the enhancement of entrepreneurship within the tea production supply chain, especially, within stakeholders who deliver digital transformation support services.

Introduction: Today the ability of nations to develop and implement innovations is core for their international competitiveness. Ukraine is striving for innovation progress; however, its innovation performance is relatively low. The research problem is to find the bottlenecks, affecting Ukraine’s innovation capability.

Purpose: This study aims to research the national innovation capability profiles, based on cluster analysis, to develop an understanding of drivers and threats for the innovation capability of Ukraine.

Need of the study: The knowledge-based economy, which had already turned into one of the most efficient developmental models of the 21st century, became a key driver of international competitiveness for the leading developed countries due to their progressive structural shifts towards the growth of high-technology manufacturing and knowledge-intensive sectors. These trends are significant to capture for the sake of increasing the innovation capability of the economy of Ukraine.

Methodology: The study is based on the K-means clustering method, which is employed for identifying 10 country clusters based on the indicators of their R&D and innovation activities, which allowed us to assess the innovation capability of Ukraine in comparison with 140 countries of the world. Data selection and normalisation were based on the 2019 Global Competitiveness Report indicators.

Findings: The study showed that Ukraine’s innovation capability problems are typical for most developing countries and are prevalently connected to low R&D expenditures, patent applications, and international co-invention activities. Most countries, except for the technologically developed ones, follow the so-called ‘passive technological learning’ strategies, which usually result in low economic productivity.

Practical implications: Several innovation policy implications have been developed for the government of Ukraine based on the cluster analysis results and accounting for the problems of the national innovation system (NIS).

The environmental characteristics of international football can vary significantly from one country to another. As a result, the economic and market possibilities and the objectives of each national league are very heterogeneous. This article aims to examine the differences in revenue structures amongst European national football leagues (n = 50) and cluster them based on these structures. It also investigates which revenue structure would be more effective for similar leagues, considering the previously mentioned varying environmental characteristics of international football.

The study utilises a theoretical framework of business modelling, applied in a unique way to league organisers of national championships. Data on sports and business aspects were collected from sources such as the Union of European Football Associations (UEFA) Financial Benchmarking Reports, transfermarkt.de and related sources for the period 2015 to 2018. K-means cluster analysis, using the Euclidean distance approach, was employed to develop clusters based on revenue sources over a four-year average.

The paper presents the characteristics and year-to-year changes of nine developed clusters. Throughout the analysis, variables such as average overpayment and inequality between player values amongst leagues were prioritised. The study's practical implications can assist league organisers in enhancing the competitiveness of their leagues, supported by short case studies that provide illustrative examples.

The novelty of the current article lies in introducing innovative variables such as the variance of player value whilst focussing on meso-level analysis, providing a fresh contribution to the existing literature in the field for understanding revenue structures and performance in European national football leagues.

Machine learning (ML), and deep learning in particular, is gaining traction across a myriad of real-life applications. Portfolio management is no exception. This paper provides a systematic literature review of deep learning applications for portfolio management. The findings are likely to be valuable for industry practitioners and researchers alike, experimenting with novel portfolio management approaches and furthering investment management practice.

This review follows the guidance and methodology of Linnenluecke et al. (2020), Massaro et al. (2016) and Fisch and Block (2018) to first identify relevant literature based on an appropriately developed search phrase, filter the resultant set of publications and present descriptive and analytical findings of the research itself and its metadata.

The authors find a strong dominance of reinforcement learning algorithms applied to the field, given their through-time portfolio management capabilities. Other well-known deep learning models, such as convolutional neural network (CNN) and recurrent neural network (RNN) and its derivatives, have shown to be well-suited for time-series forecasting. Most recently, the number of papers published in the field has been increasing, potentially driven by computational advances, hardware accessibility and data availability. The review shows several promising applications and identifies future research opportunities, including better balance on the risk-reward spectrum, novel ways to reduce data dimensionality and pre-process the inputs, stronger focus on direct weights generation, novel deep learning architectures and consistent data choices.

Several systematic reviews have been conducted with a broader focus of ML applications in finance. However, to the best of the authors’ knowledge, this is the first review to focus on deep learning architectures and their applications in the investment portfolio management problem. The review also presents a novel universal taxonomy of models used.

Understanding consumer brand relationships from the perspective of the consumer has been a research topic for years. Despite this, there are still various ways in which the construct is interpreted. This paper aims to identify the most typical interpretation of brand relationships by consumers.

A four-study prototype analysis was conducted, in which a bottom-up approach was applied to identify lay people’s conceptualization of consumer brand relationships.

The prototype analysis generates a comprehensive list of features of consumer brand relationships that provide a nuanced understanding of the concept. The most typical characteristics of a brand relationship according to consumers are quality, bond, value and joy. Comparing this relationship prototype with existing literature shows that there may be a gap between theory and practice regarding the concept of brand relationship.

The prototypical conceptualization of brand relationships shows which aspects play a role in consumers' most common interpretation of the construct. This provides an opportunity to assess the validity of existing conceptualizations of brand relationships. Knowing which aspects are most relevant for consumers’ brand relationships allows brands to make adjustments as needed and improve at establishing and maintaining relationships with consumers.

The overall goal of the data mining process is to extract information from an extensive data set and make it understandable for further use. When working with large volumes of unstructured data in research information systems, it is necessary to divide the information into logical groupings after examining their quality before attempting to analyze it. On the other hand, data quality results are valuable resources for defining quality excellence programs of any information system. Hence, the purpose of this study is to discover and extract knowledge to evaluate and improve data quality in research information systems.

Clustering in data analysis and exploiting the outputs allows practitioners to gain an in-depth and extensive look at their information to form some logical structures based on what they have found. In this study, data extracted from an information system are used in the first stage. Then, the data quality results are classified into an organized structure based on data quality dimension standards. Next, clustering algorithms (K-Means), density-based clustering (density-based spatial clustering of applications with noise [DBSCAN]) and hierarchical clustering (balanced iterative reducing and clustering using hierarchies [BIRCH]) are applied to compare and find the most appropriate clustering algorithms in the research information system.

This paper showed that quality control results of an information system could be categorized through well-known data quality dimensions, including precision, accuracy, completeness, consistency, reputation and timeliness. Furthermore, among different well-known clustering approaches, the BIRCH algorithm of hierarchical clustering methods performs better in data clustering and gives the highest silhouette coefficient value. Next in line is the DBSCAN method, which performs better than the K-Means method.

In the data quality assessment process, the discrepancies identified and the lack of proper classification for inconsistent data have led to unstructured reports, making the statistical analysis of qualitative metadata problems difficult and thus impossible to root out the observed errors. Therefore, in this study, the evaluation results of data quality have been categorized into various data quality dimensions, based on which multiple analyses have been performed in the form of data mining methods.

Although several pieces of research have been conducted to assess data quality results of research information systems, knowledge extraction from obtained data quality scores is a crucial work that has rarely been studied in the literature. Besides, clustering in data quality analysis and exploiting the outputs allows practitioners to gain an in-depth and extensive look at their information to form some logical structures based on what they have found.

The common methods for clustering time series are the use of specific distance criteria or the use of standard clustering algorithms. Ensemble clustering is one of the common techniques used in data mining to increase the accuracy of clustering. In this study, based on segmentation, selecting the best segments, and using ensemble clustering for selected segments, a multistep approach has been developed for the whole clustering of time series data.

First, this approach divides the time series dataset into equal segments. In the next step, using one or more internal clustering criteria, the best segments are selected, and then the selected segments are combined for final clustering. By using a loop and how to select the best segments for the final clustering (using one criterion or several criteria simultaneously), two algorithms have been developed in different settings. A logarithmic relationship limits the number of segments created in the loop.

According to Rand's external criteria and statistical tests, at first, the best setting of the two developed algorithms has been selected. Then this setting has been compared to different algorithms in the literature on clustering accuracy and execution time. The obtained results indicate more accuracy and less execution time for the proposed approach.

This paper proposed a fast and accurate approach for time series clustering in three main steps. This is the first work that uses a combination of segmentation and ensemble clustering. More accuracy and less execution time are the remarkable achievements of this study.

For privacy protection, federated learning based on data separation allows machine learning models to be trained on remote devices or in isolated data devices. However, due to the limited resources such as bandwidth and power of local devices, communication in federated learning can be much slower than in local computing. This study aims to improve communication efficiency by reducing the number of communication rounds and the size of information transmitted in each round.

This paper allows each user node to perform multiple local trainings, then upload the local model parameters to a central server. The central server updates the global model parameters by weighted averaging the parameter information. Based on this aggregation, user nodes first cluster the parameter information to be uploaded and then replace each value with the mean value of its cluster. Considering the asymmetry of the federated learning framework, adaptively select the optimal number of clusters required to compress the model information.

While maintaining the loss convergence rate similar to that of federated averaging, the test accuracy did not decrease significantly.

By compressing uplink traffic, the work can improve communication efficiency on dynamic networks with limited resources.