Browsing by Author "Yusuff, S. R."

Now showing 1 - 5 of 5
  • Thumbnail Image
    Item
    A Prediction Model for Bitcoin Cryptocurrency Prices
    (Springer, Cham, 2022-04-20) Aorowolo, M. O.; Ayegba, P.; Yusuff, S. R.; Misra, S
    Cryptocurrencies like Bitcoin are a contentious and difficult technological innovation in today's financial system. With huge improvements in financial markets, machine learning and Artificial Intelligence aided trading have piqued interest in recent years. This study suggests a predicting model for blockchain bitcoin cryptocurrency prices and its profitability trading strategies using machine learning algorithms (ICA-Firefly and SVMs). For the prediction analysis of Bitcoin cryptocurrency data, this study combines ICA-Firefly with SVM algorithms. The model was tested on a large dataset of 2,194 samples, and its performance was analyzed in terms of evaluation metrics. In evaluation to state of the art, the ICAFirefly with L-SVM and Sigmoid SVM classification approach performs well on the bitcoin sample dataset, with an accuracy of 95% and 97%, respectively. The ICA-Firefly with the SVM model can be adopted as a viable financial system sustainability management strategy.
  • Thumbnail Image
    Item
    A Study of Impacts of Artificial Intelligence on COVID-19 Prediction, Diagnosis, Treatment, and Prognosis
    (An International Pan-African Multidisciplinary Journal of the SMART Research Group International Centre for IT & Development (ICITD) USA © Creative Research Publishers, 2022-12-04) Isiaka, R. M.; Babatunde, R. S.; Ajao, J. F.; Yusuff, S. R.; Popoola, D. D.; Arowolo, M. O.; Adewole, K. S.
    Following the identification of Coronavirus Disease 2019 (COVID-19) in Wuhan, China in December 2019, AI researchers have teamed up with a health specialist to combat the virus. This study explores the medical and non-medical areas of COVID-19 that AI has impacted: the prevalence of the AI technologies adopted across all stages of the pandemic, the collaboration networks of global AI researchers, and the open issues. 21,219 papers from ACM Digital, Science Direct and Google Scholar were examined. Adherence to the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) framework and utilizing the PICO (population, intervention, comparison, and outcome) paradigm, guided the inclusion of researches in the review. Tables and graphs were utilized to display the results. Analysis revealed that AI has impacted 4 molecular, 4 clinical, and 7 societal areas of COVID-19. Deep Learning among other AI technologies was traced to all aspects of the pandemic. 2173 authors and co-authors were traced to these achievements, while 32 of the most connected 51 authors were affiliated with institutions in China, 18 to the United States, and 1 to Europe. The open issues identified had to do with the quality of datasets, AI model deployment, and privacy issues. This study demonstrates how AI may be utilized for COVID-19 diagnosis, prediction, medication and vaccine identification, prognosis, and contact person monitoring. This investigation began at the beginning of the epidemic and continued until the first batch of vaccinations received approval. The study provided collaboration opportunities for AI researchers and revealed open issues that will spike further research toward preparing the world for any future pandemic
  • Thumbnail Image
    Item
    An Evaluation of performance of k-NN and Mahalanobis Distance on Local Binary Pattern Feature representation for Access Control in Face Recognition System.
    (The Creative Research Networks for The African Society for Information & Communication Technology, USA, 2017-03) Yusuff, S. R.; Balogun, F. B.; Babtunde A. N.; Babatunde, R. S.
    A vast majority of the existing face recognition techniques encounter misclassifications when there is a large variation of the subject in terms of pose, illumination and expression as well as artefact such as occlusion. With the usage of Local Binary Pattern (LBP), local feature representation which preserves local primitives and texture information, robust against illumination, expression and occlusion was realized in the feature extraction process. Mahalanobis distance and k-NN classifier were employed for recognition as basis for comparison. An assessment of both algorithms was considered. It was observed that Mahalanobis distance measure outperformed the k-NN classifier for access control and recognition system due to its high recognition rate (98.4%), genuine acceptance rate (98.37) and its stern adherence to both FAR and FRR (0.0144, 1.568) than kNN with 94.9% and (0.392 and 2.895) respectively. This result demonstrates the robustness of Mahalanobis distance classifier for use access control system
  • Thumbnail Image
    Item
    Offline yoruba word recognition system based on capsule neural network
    (Ilorin Journal of Computer Science and Information Technology, 2022-06-30) Ajao, J. F.; Olorunmaiye, E; Mabayoje, M. A.; Yusuff, S. R.; Bajeh, A.
    The need for Digital Document archiving has resulted in several efforts in Image Document Analysis, where scanned historical/ indigenous documents can be stored in digitized form and made readily accessible and available for future use. Visual system in human uses an highly space variant in sampling, coding, processing and understanding of variations in human handwritings, which possess lots of challenges or difficulty for machine to understand. Consequently, it is important to preserve our indigenous heritage values by way of digitization. In this regard, the research work models a Yoruba Script Recognition System using Capsnet, based on the peculiar features of Yoruba alphabet. Yoruba bird names were gathered from the Yoruba dictionary and were written by 20 indigenous literate writers in 50 different writing styles. This was used to model HYSR system. The training time and testing after training different epoch produces 89.2222 and the testing time generated is 0.8667. The model accuracy gives 85.49%, which shows a promising result.
  • Thumbnail Image
    Item
    YORÙBÁ CHARACTER RECOGNITION SYSTEM USING CONVOLUTIONAL RECURRENT NEURAL NETWORK
    (Ugur SEN, 2022-09-01) Ajao. J. F.; Yusuff, S. R.; Ajao, A. O.
    Abstract: Handwritten recognition systems enable automatic recognition of human handwritings, thereby increasing humancomputer interaction. Despite enormous efforts in handwritten recognition, little progress has been made due to the variability of human handwriting, which presents numerous difficulties for machines to recognize. It was discovered that while tremendous progress has been made in handwritten recognition of English and Arabic languages, very little work has been done on Yorùbá handwritten characters. Those few works, in turn, made use of Hidden Markov Model (HMM), Support Vector Machine (SVM), Bayes theorem, and decision tree algorithms. To integrate and save one of Nigeria's indigenous languages from extinction, as well as to make Yorùbá documents accessible and available in the digital world, this research work was undertaken. The research presents a convolutional recurrent neural network (CRNN) for the recognition of Yorùbá handwritten characters. Data were collected from students of Kwara State University who were literate Yorùbá writers. The collected data were subjected to some level of preprocessing such as grayscale, binarization, and normalization in order to remove perturbations introduced during the digitization process. The convolutional recurrent neural network model was trained using the preprocessed images. The evaluation was conducted using the acquired Yorùbá characters, 87.5% of the acquired images were used for the training while 12.5% were used to evaluate the developed system. As there is currently no publicly available database of Yorùbá characters for validating Yorùbá recognition systems. The resulting recognition accuracy was 87.2% while the characters with under dot and diacritic signs has low recognition accuracy