Publicação
Overcoming over–indebtedness with AI - A case study on the application of AutoML to research
| dc.contributor.advisor | Castelli, Mauro | |
| dc.contributor.author | Costa, Victor Cardoso Reis | |
| dc.date.accessioned | 2021-04-08T10:56:13Z | |
| dc.date.available | 2021-04-08T10:56:13Z | |
| dc.date.issued | 2021-03-30 | |
| dc.description | Dissertation presented as the partial requirement for obtaining a Master's degree in Data Science and Advanced Analytics | pt_PT |
| dc.description.abstract | This research examines how artificial intelligence may contribute to better understanding and overcoming over-indebtedness in contexts of high poverty risk. This study uses a field database of 1,654 over-indebted households to identify distinguishable clusters and to predict its risk factors. First, unsupervised machine learning generated three overindebtedness clusters: low-income (31.27%), low credit control (37.40%), and crisis-affected households (31.33%). These served as basis for a better understanding on the complex issue that is over-indebtedness. Second, a predictive model was developed to serve as a tool for policymakers and advisory services by streamlining the classification of overindebtedness profiles. On building such model, an AutoML approach was leveraged achieving performant results (92.1% accuracy score). Furthermore, within the AutoML framework, two techniques were employed, leading to a deeper discussion on the benefits and inner workings of such strategy. Ultimately, this research looks to contribute on three fronts: theoretical, by unfolding previously unexplored characteristics on the concept of over-indebtedness; methodological, by proposing AutoML as a powerful research tool accessible to investigators on many backgrounds; and social, by building real-world applications that aim at mitigating over-indebtedness and, consequently, poverty risk. | pt_PT |
| dc.identifier.tid | 202692469 | pt_PT |
| dc.identifier.uri | http://hdl.handle.net/10362/115197 | |
| dc.language.iso | eng | pt_PT |
| dc.rights.uri | http://creativecommons.org/licenses/by/4.0/ | pt_PT |
| dc.subject | over-indebtedness | pt_PT |
| dc.subject | poverty risk | pt_PT |
| dc.subject | credit control | pt_PT |
| dc.subject | artificial intelligence | pt_PT |
| dc.subject | machine learning | pt_PT |
| dc.subject | automated machine learning | pt_PT |
| dc.subject | automl | pt_PT |
| dc.title | Overcoming over–indebtedness with AI - A case study on the application of AutoML to research | pt_PT |
| dc.type | master thesis | |
| dspace.entity.type | Publication | |
| rcaap.rights | openAccess | pt_PT |
| rcaap.type | masterThesis | pt_PT |
| thesis.degree.name | Mestrado em Métodos Analíticos Avançados | pt_PT |
Ficheiros
Principais
1 - 1 de 1