Department of Mathematical and Computational Sciences

Permanent URI for this community

https://univendspace.univen.ac.za/handle/11602/1927

Browse

Browsing Department of Mathematical and Computational Sciences by Author "Chibaya, Colin"

Now showing 1 - 2 of 2
  • Thumbnail Image
    ItemOpen Access
    A Neural Network Enhanced RSA Model Towards a Confidentiality-Integrity-Authenticity Compliant Hybrid
    (2021-03-19) Magoma, Promise Tshepiso; Chibaya, Colin; Mathebula, Daphney
    Data security is an important aspect in the field of data science where data collection, analysis, interpretation, and sharing are a primary goal. To prevent unauthorized access to data, creative methods to securing data are sought. Cryptography is about the development of algorithms with which to hide data. The three key objectives of cryptography are to achieve data confidentiality (C), data integrity (I), and data authenticity (A). Algorithms that can achieve all these three objectives at once are said to be CIA compliant. However, there are barely any algorithms out there that can satisfy these three objectives in one goal. However, CIA-compliant cryptosystems are, to the best of our knowledge, rare. The RSA algorithm is a compelling cryptosystem that was mainly designed to achieve data confidentiality. It demonstrates attractive properties for improvement towards CIA compliancy. Some research has tried to upgrade the RSA algorithm by combining it with the DH model or the El Gamal model. However, still, the outcome would either be CI or CA compliant, leaving out one of the three objectives. This study investigates the improvement of the RSA algorithm by incorporating a neural network to learn data integrity and data authenticity towards creating a CIA-compliant hybrid RSA model. To the best of our knowledge, this is the first time a neural network has been proposed for improving the RSA model towards CIA compliance. Experimental results indicate that a neural network can learn data integrity and data authenticity in RSA encrypted messages. Data analysis affirmed that neural network learning can be generalized. A conclusion that the RSA algorithm can be upgraded towards CIA compliance when a neural network is incorporated was arrived at. These findings have implications for the commercial standing of the RSA algorithm as well as for the body of knowledge in the cryptography domain.
  • Thumbnail Image
    ItemOpen Access
    Short term wind power forecasting in South Africa using neural networks
    (2020-08-11) Daniel, Lucky Oghenechodja; Sigauke, Caston; Chibaya, Colin; Mbuvha, Rendani
    Wind offers an environmentally sustainable energy resource that has seen increasing global adoption in recent years. However, its intermittent, unstable and stochastic nature hampers its representation among other renewable energy sources. This work addresses the forecasting of wind speed, a primary input needed for wind energy generation, using data obtained from the South African Wind Atlas Project. Forecasting is carried out on a two days ahead time horizon. We investigate the predictive performance of artificial neural networks (ANN) trained with Bayesian regularisation, decision trees based stochastic gradient boosting (SGB) and generalised additive models (GAMs). The results of the comparative analysis suggest that ANN displays superior predictive performance based on root mean square error (RMSE). In contrast, SGB shows outperformance in terms of mean average error (MAE) and the related mean average percentage error (MAPE). A further comparison of two forecast combination methods involving the linear and additive quantile regression averaging show the latter forecast combination method as yielding lower prediction accuracy. The additive quantile regression averaging based prediction intervals also show outperformance in terms of validity, reliability, quality and accuracy. Interval combination methods show the median method as better than its pure average counterpart. Point forecasts combination and interval forecasting methods are found to improve forecast performance.