A framework to select a classification algorithm in electricity fraud detection

Pazi, Sisa; Clohessy, Chantelle M.; Sharp, Gary D.

doi:10.17159/sajs.2020/8189

Cited by 5 publications

(3 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The sample with the new scores was used to cross-validate the results. Cross-validation was repeated 500 times using the technique described by Pazi et al [11] Lastly, the four models were compared by analysing the results from 500 validation samples.…”

Section: Methodsmentioning

confidence: 99%

Prediction of in-hospital mortality: An adaptive severity- of-illness score for a tertiary ICU in South Africa

Pazi

Sharp

Merwe

2022

SAJCC

View full text Add to dashboard Cite

Background. A scoring system based on physiological conditions was developed in 1984 to assess the severity of illness. This version, and subsequent versions, were labelled Simplified Acute Physiology Scores (SAPS). Each extension addressed limitations in the earlier version, with the SAPS III model using a data-driven approach. However, the SAPS III model did not include data collected from the African continent, thereby limiting the generalisation of the results. Objectives. To propose a scoring system for assessing severity of illness at intensive care unit (ICU) admission and a model for prediction of in-hospital mortality, based on the severity of illness score. Methods. This is a prospective cohort study which included patients who were admitted to an ICU in a South African tertiary hospital in 2017. Logistic regression modelling was used to develop the proposed scoring system, and the proposed mortality prediction model. Results. The study included 829 patients. Less than a quarter of patients (21.35%;n=177) died during the study period. The proposed model exhibited good calibration and excellent discrimination. Conclusion. The proposed scoring system is able to assess severity of illness at ICU admission, while the proposed statistical model may be used in the prediction of in-hospital mortality.

show abstract

Section: Methodsmentioning

confidence: 99%

Prediction of in-hospital mortality: An adaptive severity- of-illness score for a tertiary ICU in South Africa

Pazi

Sharp

Merwe

2022

SAJCC

View full text Add to dashboard Cite

show abstract

“…This approach can be used for classifications in another aspect of machine learning applications because of its low computational cost in removing irrelevant power consumption patterns example includes principal component analysis (PCA) in [24], linear discriminant analysis (LDA) in [36] and quadratic discriminant analysis (QDA) in [1,8].…”

Section: Unsupervised Detection Models-dimensionality Reduction Algor...mentioning

confidence: 99%

Systematic literature review of the techniques for household electrical appliance anomaly detections and knowledge extractions

Rauf

Adekoya

2023

Journal of Electrical Systems and Inf Technol

View full text Add to dashboard Cite

The demand for electricity at home has increased in recent times globally, this high demand for continuous, stable and affordable power can be attributed to the demand for comfortable lifestyle of consumers but the quality and efficiency of the appliances being used remain questionable. Malfunctioning appliances usually show a power signature statistically different from their normal behavior, which can lead to higher energy consumption or more serious damages. As a result, numerous studies in recent times have been conducted on the household electrical appliance anomaly behaviors to find the root-cause of these anomalies using machine learning techniques and algorithms. This study attempted to undertake a systematic and critical review of ninety-two (92) research works reported in academic journals over fifteen (15) years (2006–2021) in the area of household electrical appliance anomaly detections and knowledge extraction using machine learning. The various techniques used in these reports were clustered based on machine learning-based techniques, statistical techniques and physical based approach techniques and the parameters adopted, such as machine learning algorithms, feature extraction approaches, anomaly detection levels, computing platforms and application scenarios. This clustering was done based on the following criteria: the nature of a dataset and the number of data sources used, the data timeframe, the machine learning algorithms used, machine learning task, used accuracy and error metrics and software packages used for modeling. For the number of data source used, the results revealed that 81.2% of documents reviewed used single sources and Autoregressive integrated moving average (ARIMA) was the highest implemented regression model (60.9%), the probability model that was mostly implemented was the Bayesian network. Furthermore, the study revealed that, root-mean-square error (RMSE) accounted 35% was the most used error metric among household appliance abnormal behaviors, followed by mean absolute percentage error MAPE which accounted 32%. The study further revealed that 46% of appliance abnormal detections was based on weather parameters, and historical energy consumption. Finally, we recap the challenges and limitations for further research in electrical appliance anomaly detections locally and globally.

show abstract

“…Adicionalmente, [13] realizaron una investigación para predecir la demanda de energía eléctrica de edificios, utilizando un predictor basado en el algoritmo de K vecinos más cercanos, que resultó significativamente más preciso que otros modelos utilizados previamente. En su estudio, [14] diseñaron una estructura para seleccionar el mejor algoritmo de clasificación para detectar fraudes en el consumo de energía eléctrica. Utilizaron distintas métricas para comparar los algoritmos, entre los cuales estuvo K vecinos más cercanos.…”

Section: Introductionunclassified

Uso de algoritmos de aprendizaje automático para analizar datos de energía eléctrica facturada. Caso: Chile 2015 – 2021

Ramírez

2022

I+D Tecnológico

View full text Add to dashboard Cite

En el mercado eléctrico chileno, los usuarios finales se clasifican en clientes libres y clientes regulados. El análisis del consumo energético de los clientes regulados es importante para efectos del diseño y aplicación de las políticas públicas del sector. En esta investigación se hace el análisis de los datos de energía eléctrica facturada mensual de los clientes regulados de Chile, durante el período 2015-2021, con el fin de detectar patrones y predecir la categoría a la que pertenecen. Se utilizan los algoritmos K-Means para la detección de patrones, K-NN para la predicción de la categoría de los clientes, y PCA para determinar las variables más significativas dentro del conjunto de datos. Con K-Means se encontró que los datos se agrupan de acuerdo con el tipo de cliente, con K-NN se obtuvo un modelo que permite predecir a qué tipo de clientes pertenecen los datos, y con PCA se encontró que las variables tipo de cliente, el año y el mes, son las más importantes en el conjunto de datos. Más del 96% de los clientes analizados corresponde al tipo residencial, quienes consumieron el 50% de la energía facturada durante el período de estudio, y además imponen la estacionalidad mensual de los datos. Los resultados obtenidos son de ayuda para el establecimiento y revisión de las políticas aplicadas a los clientes regulados, en cuanto a tarifas, límites de consumo en invierno, y eficiencia energética. Se recomienda continuar la investigación orientándola hacia la predicción del consumo de energía eléctrica.

show abstract

A framework to select a classification algorithm in electricity fraud detection

Cited by 5 publications

References 19 publications

Prediction of in-hospital mortality: An adaptive severity- of-illness score for a tertiary ICU in South Africa

Prediction of in-hospital mortality: An adaptive severity- of-illness score for a tertiary ICU in South Africa

Systematic literature review of the techniques for household electrical appliance anomaly detections and knowledge extractions

Uso de algoritmos de aprendizaje automático para analizar datos de energía eléctrica facturada. Caso: Chile 2015 – 2021

Contact Info

Product

Resources

About