Cargando…
Experimental Application of Machine Learning on Financial Inclusion Data for Governance in Eswatini
An objectives of good governance is to increase capital base of small scale businesses (SSB) in order to encourage more investments and hence increase employment rate. Embracing good financial inclusion (FI) schemes in a country helps to ensure that entrepreneurs of SSB have access to financial serv...
Autores principales: | , , , |
---|---|
Formato: | Online Artículo Texto |
Lenguaje: | English |
Publicado: |
2020
|
Materias: | |
Acceso en línea: | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7134229/ http://dx.doi.org/10.1007/978-3-030-45002-1_36 |
_version_ | 1783517795454025728 |
---|---|
author | Akinnuwesi, Boluwaji A. Fashoto, Stephen G. Metfula, Andile S. Akinnuwesi, Adetutu N. |
author_facet | Akinnuwesi, Boluwaji A. Fashoto, Stephen G. Metfula, Andile S. Akinnuwesi, Adetutu N. |
author_sort | Akinnuwesi, Boluwaji A. |
collection | PubMed |
description | An objectives of good governance is to increase capital base of small scale businesses (SSB) in order to encourage more investments and hence increase employment rate. Embracing good financial inclusion (FI) schemes in a country helps to ensure that entrepreneurs of SSB have access to financial services and hence meet their needs. In this paper we studied FI scheme in Kingdom of Eswatini with the view to establish the extent to which SSB have access to funds in running their businesses such that they could satisfy the target population and meet their desired goals. We got FI dataset for Eswatini for 2018 from Finscope database. Finscope 2018 dataset contains 1385 attributes with 2928 records. This study extracted attributes based on payment channel, registered/unregistered business, usage of commercial banks/insurance/mobile money and source of income for households from the Finscope database. We identified lot of missing data and hence replaced them using Mode method of preprocessing module in WEKA. We split the datasets and carried out cross validation on it. Training data is 80% of the datasets and 20% was used for testing. We carefully classified FI for selected parameters for Hhohho, Manzini, Shiselweni and Lubombo regions of Eswatini using Logistic regression with 80% for training and 10 fold cross-validation. The best 10 fold cross-validation recall rate for Manzini region using support vector machine (SVM) is 69.4% and 63.4% using logistic regression. These results show that veracity of FI dataset is weak and this is due to large number of missing data. |
format | Online Article Text |
id | pubmed-7134229 |
institution | National Center for Biotechnology Information |
language | English |
publishDate | 2020 |
record_format | MEDLINE/PubMed |
spelling | pubmed-71342292020-04-06 Experimental Application of Machine Learning on Financial Inclusion Data for Governance in Eswatini Akinnuwesi, Boluwaji A. Fashoto, Stephen G. Metfula, Andile S. Akinnuwesi, Adetutu N. Responsible Design, Implementation and Use of Information and Communication Technology Article An objectives of good governance is to increase capital base of small scale businesses (SSB) in order to encourage more investments and hence increase employment rate. Embracing good financial inclusion (FI) schemes in a country helps to ensure that entrepreneurs of SSB have access to financial services and hence meet their needs. In this paper we studied FI scheme in Kingdom of Eswatini with the view to establish the extent to which SSB have access to funds in running their businesses such that they could satisfy the target population and meet their desired goals. We got FI dataset for Eswatini for 2018 from Finscope database. Finscope 2018 dataset contains 1385 attributes with 2928 records. This study extracted attributes based on payment channel, registered/unregistered business, usage of commercial banks/insurance/mobile money and source of income for households from the Finscope database. We identified lot of missing data and hence replaced them using Mode method of preprocessing module in WEKA. We split the datasets and carried out cross validation on it. Training data is 80% of the datasets and 20% was used for testing. We carefully classified FI for selected parameters for Hhohho, Manzini, Shiselweni and Lubombo regions of Eswatini using Logistic regression with 80% for training and 10 fold cross-validation. The best 10 fold cross-validation recall rate for Manzini region using support vector machine (SVM) is 69.4% and 63.4% using logistic regression. These results show that veracity of FI dataset is weak and this is due to large number of missing data. 2020-03-10 /pmc/articles/PMC7134229/ http://dx.doi.org/10.1007/978-3-030-45002-1_36 Text en © IFIP International Federation for Information Processing 2020 This article is made available via the PMC Open Access Subset for unrestricted research re-use and secondary analysis in any form or by any means with acknowledgement of the original source. These permissions are granted for the duration of the World Health Organization (WHO) declaration of COVID-19 as a global pandemic. |
spellingShingle | Article Akinnuwesi, Boluwaji A. Fashoto, Stephen G. Metfula, Andile S. Akinnuwesi, Adetutu N. Experimental Application of Machine Learning on Financial Inclusion Data for Governance in Eswatini |
title | Experimental Application of Machine Learning on Financial Inclusion Data for Governance in Eswatini |
title_full | Experimental Application of Machine Learning on Financial Inclusion Data for Governance in Eswatini |
title_fullStr | Experimental Application of Machine Learning on Financial Inclusion Data for Governance in Eswatini |
title_full_unstemmed | Experimental Application of Machine Learning on Financial Inclusion Data for Governance in Eswatini |
title_short | Experimental Application of Machine Learning on Financial Inclusion Data for Governance in Eswatini |
title_sort | experimental application of machine learning on financial inclusion data for governance in eswatini |
topic | Article |
url | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7134229/ http://dx.doi.org/10.1007/978-3-030-45002-1_36 |
work_keys_str_mv | AT akinnuwesiboluwajia experimentalapplicationofmachinelearningonfinancialinclusiondataforgovernanceineswatini AT fashotostepheng experimentalapplicationofmachinelearningonfinancialinclusiondataforgovernanceineswatini AT metfulaandiles experimentalapplicationofmachinelearningonfinancialinclusiondataforgovernanceineswatini AT akinnuwesiadetutun experimentalapplicationofmachinelearningonfinancialinclusiondataforgovernanceineswatini |