Browsing by Author "Afriyie, Jonathan Kwaku"

Now showing 1 - 3 of 3
  • No Thumbnail Available
    Item
    A hybrid forecasting technique for infection and death from the mpox virus.
    (Digital Health, 2023) Iftikhar, Hasnain; Daniyal, Muhammad; Quresh, Moiz; Tawaiah, Kassim; Ansah, Richard Kwame; Afriyie, Jonathan Kwaku; 0000-0001-6997-7969
    Objectives: The rising of new cases and death counts from the mpox virus (MPV) is alarming. In order to mitigate the impact of the MPV it is essential to have information of the virus’s future position using more precise time series and stochastic models. In this present study, a hybrid forecasting system has been developed for new cases and death counts for MPV infection using the world daily cumulative confirmed and death series. Methods: The original cumulative series was decomposed into new two subseries, such as a trend component and a stochastic series using the Hodrick–Prescott filter. To assess the efficacy of the proposed models, a comparative analysis with several widely recognized benchmark models, including auto-regressive (AR) model, auto-regressive moving average (ARMA) model, non-parametric auto-regressive (NPAR) model and artificial neural network (ANN), was performed. Results: The introduction of two novel hybrid models, HPF1 1 and HPF4 3, which demonstrated superior performance compared to all other models, as evidenced by their remarkable results in key performance indicators such as root mean square error (RMSE), mean absolute error (MAE) and mean absolute percentage error (MAPE), is a significant advancement in disease prediction. Conclusion: The new models developed can be implemented in forecasting other diseases in the future. To address the current situation effectively, governments and stakeholders must implement significant changes to ensure strict adherence to standard operating procedures (SOPs) by the public. Given the anticipated continuation of increasing trends in the coming days, these measures are essential for mitigating the impact of the outbreak.
  • No Thumbnail Available
    Item
    A supervised machine learning algorithm for detecting and predicting fraud in credit card transactions
    (Decision Analytics Journal, 2023-03-06) Afriyie, Jonathan Kwaku; Tawiah, Kassim; Pels, Wilhemina Adoma; Addai-Henne, Sandra; Dwamena, Harriet Achiaa; Owiredu, Emmanuel Odame; Ayeh, Samuel Amening; Eshun, John; 0000-0001-6997-7969
    Fraudsters are now more active in their attacks on credit card transactions than ever before. With the advancement in data science and machine learning, various algorithms have been developed to determine whether a transaction is fraudulent. We study the performance of three different machine learning models: logistic regression, random forest, and decision trees to classify, predict, and detect fraudulent credit card transactions. We compare these models’ performance and show that random forest produces a maximum accuracy of 96% (with an area under the curve value of 98.9%) in predicting and detecting fraudulent credit card transactions. Thus, we recommend random forest as the most appropriate machine learning algorithm for predicting and detecting fraud in credit card transactions. Credit Card holders above 60 years were found to be mostly victims of these fraudulent transactions, with a greater proportion of fraudulent transactions occurring between the hours of 22:00GMT and 4:00GMT.
  • No Thumbnail Available
    Item
    A supervised machine learning algorithm for detecting and predicting fraud in credit card transactions
    (Decision Analytics Journal, 2023-03) Afriyie, Jonathan Kwaku; Tawiah, Kassim; Pels, Wilhemina Adoma; Addai-Henne, Sandra; Dwamena, Harriet Achiaa; Owiredu, Emmanuel Odame; Ayeh, Samuel Amening; John Eshun; https://orcid.org/0000-0001-7881-3069
    Fraudsters are now more active in their attacks on credit card transactions than ever before. With the advancement in data science and machine learning, various algorithms have been developed to determine whether a transaction is fraudulent. We study the performance of three different machine learning models: logistic regression, random forest, and decision trees to classify, predict, and detect fraudulent credit card transactions. We compare these models’ performance and show that random forest produces a maximum accuracy of 96% (with an area under the curve value of 98.9%) in predicting and detecting fraudulent credit card transactions. Thus, we recommend random forest as the most appropriate machine learning algorithm for predicting and detecting fraud in credit card transactions. Credit Card holders above 60 years were found to be mostly victims of these fraudulent transactions, with a greater proportion of fraudulent transactions occurring between the hours of 22:00GMT and 4:00GMT.