Validation and adaptation of statistical models based on the SAPS III score to predict in-hospital mortality in a South African ICU
- Authors: Pazi, Sisa
- Date: 2023-04
- Subjects: Police -- South Africa -- Eastern Cape , Statistics – South Africa , Mortality – South Africa
- Language: English
- Type: Doctoral's theses , text
- Identifier: http://hdl.handle.net/10948/61360 , vital:70602
- Description: In-hospital mortality prediction remains an important task in Intensive Care Units (ICUs). In particular, the estimated in-hospital mortality risk is essential to describe case-mix, for research and clinical auditing purposes. Furthermore, in settings with limited hospital resources (e.g beds) such as the South African public health care system, the estimated in-hospital mortality risk is essential for resource allocation and to inform local patient triage guidelines. Commonly used models for prediction of in-hospital mortality in ICU patients includes, but not limited to, the Simplified Acute Physiology Score III (SAPS III). The SAPS III model was developed in 2005. Notably, the SAPS III model was developed without data collected from African based hospitals. Given the general application of the SAPS III model, including benchmarking and quality control, the development of such a model based on local data is of paramount importance. To this end, this study developed a model for prediction of in-hospital mortality based on data collected in a hospital in South Africa. Logistic regression modelling was used to develop the proposed mortality risk assessment model. The results indicated that the proposed model exhibited superior discrimination and classification abilities compared to the SAPS III model. Future research includes the external validation of the proposed model in different hospitals in South Africa. , Thesis (PhD) -- Faculty of Science, School of Computer Science, Mathematics, Physics and Statistics, 2023
- Full Text:
- Date Issued: 2023-04
- Authors: Pazi, Sisa
- Date: 2023-04
- Subjects: Police -- South Africa -- Eastern Cape , Statistics – South Africa , Mortality – South Africa
- Language: English
- Type: Doctoral's theses , text
- Identifier: http://hdl.handle.net/10948/61360 , vital:70602
- Description: In-hospital mortality prediction remains an important task in Intensive Care Units (ICUs). In particular, the estimated in-hospital mortality risk is essential to describe case-mix, for research and clinical auditing purposes. Furthermore, in settings with limited hospital resources (e.g beds) such as the South African public health care system, the estimated in-hospital mortality risk is essential for resource allocation and to inform local patient triage guidelines. Commonly used models for prediction of in-hospital mortality in ICU patients includes, but not limited to, the Simplified Acute Physiology Score III (SAPS III). The SAPS III model was developed in 2005. Notably, the SAPS III model was developed without data collected from African based hospitals. Given the general application of the SAPS III model, including benchmarking and quality control, the development of such a model based on local data is of paramount importance. To this end, this study developed a model for prediction of in-hospital mortality based on data collected in a hospital in South Africa. Logistic regression modelling was used to develop the proposed mortality risk assessment model. The results indicated that the proposed model exhibited superior discrimination and classification abilities compared to the SAPS III model. Future research includes the external validation of the proposed model in different hospitals in South Africa. , Thesis (PhD) -- Faculty of Science, School of Computer Science, Mathematics, Physics and Statistics, 2023
- Full Text:
- Date Issued: 2023-04
Statistical methods for the detection of non-technical losses: a case study for the Nelson Mandela Bay Municipality
- Authors: Pazi, Sisa
- Date: 2017
- Subjects: Nonparametric statistics Mathematical statistics , Statistics
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10948/19706 , vital:28939
- Description: Electricity is one of the most stolen commodities in the world. Electricity theft can be defined as the criminal act of stealing electrical power. Several types of electricity theft exist, including illegal connections and bypassing and tampering with energy meters. The negative financial impacts, due to lost revenue, of electricity theft are far reaching and affect both developing and developed countries. . Here in South Africa, Eskom loses over R2 Billion annually due to electricity theft. Data mining and nonparametric statistical methods have been used to detect fraudulent usage of electricity by assessing abnormalities and abrupt changes in kilowatt hour (kWh) consumption patterns. Identifying effective measures to detect fraudulent electricity usage is an active area of research in the electrical domain. In this study, Support Vector Machines (SVM), Naïve Bayes (NB) and k-Nearest Neighbour (KNN) algorithms were used to design and propose an electricity fraud detection model. Using the Nelson Mandela Bay Municipality as a case study, three classifiers were built with SVM, NB and KNN algorithms. The performance of these classifiers were evaluated and compared.
- Full Text:
- Date Issued: 2017
- Authors: Pazi, Sisa
- Date: 2017
- Subjects: Nonparametric statistics Mathematical statistics , Statistics
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: http://hdl.handle.net/10948/19706 , vital:28939
- Description: Electricity is one of the most stolen commodities in the world. Electricity theft can be defined as the criminal act of stealing electrical power. Several types of electricity theft exist, including illegal connections and bypassing and tampering with energy meters. The negative financial impacts, due to lost revenue, of electricity theft are far reaching and affect both developing and developed countries. . Here in South Africa, Eskom loses over R2 Billion annually due to electricity theft. Data mining and nonparametric statistical methods have been used to detect fraudulent usage of electricity by assessing abnormalities and abrupt changes in kilowatt hour (kWh) consumption patterns. Identifying effective measures to detect fraudulent electricity usage is an active area of research in the electrical domain. In this study, Support Vector Machines (SVM), Naïve Bayes (NB) and k-Nearest Neighbour (KNN) algorithms were used to design and propose an electricity fraud detection model. Using the Nelson Mandela Bay Municipality as a case study, three classifiers were built with SVM, NB and KNN algorithms. The performance of these classifiers were evaluated and compared.
- Full Text:
- Date Issued: 2017
- «
- ‹
- 1
- ›
- »