Depatment of Computer Science

Permanent URI for this community

https://kwasuspace.kwasu.edu.ng/handle/123456789/47

Browse

Now showing 1 - 20 of 69

Development of an Intrusion Detection System in a Computer Network.
(International Journal of Computers & Technology., 2014-03-14) Babatunde, R. S., Adewole K. S., Abdulsalam S. O. and Isiaka R. M. (2014)
The development of network technologies and application has promoted network attack both in number and severity. The last few years have seen a dramatic increase in the number of attacks, hence, intrusion detection has become the mainstream of information assurance. A computer network system should provide confidentiality, integrity and assurance against denial of service. While firewalls do provide some protection, they do not provide full protection. This is because not all access to the network occurs through the firewall. This is why firewalls need to be complemented by an intrusion detection system (IDS).An IDS does not usually take preventive measures when an attack is detected; it is a reactive rather than proactive agent. It plays the role of an informant rather than a police officer. In this research, an intrusion detection system that can be used to deny illegitimate access to some operations was developed. The IDS also controls the kind of operations performed by users (i.e. clients) on the network. However, unlike other methods, this requires no encryption or cryptographic processing on a per-packet basis. Instead, it scans the various messages sent on a network by the user. The system was developed using MicrosoftVisual Basic.
Development of Fingerprint Biometric Attendance System for Non-Academic Staff in a Tertiary Institution.
(Computer Engineering and Intelligent Systems., 2014-04-13) Adewole, K. S., Abdulsalam S. O., Babatunde R. S., Shittu T. M. and Oloyede M. O. (2014)
Institutions, companies and organisations where security and net productivity is vital, access to certain areas must be controlled and monitored through an automated system of attendance. Managing people is a difficult task for most of the organizations and maintaining the attendance record is an important factor in people management. When considering the academic institute, taking the attendance of non-academic staff on daily basis and maintaining the records is a major task. Manually taking attendance and maintaining it for a long time adds to the difficulty of this task as well as wastes a lot of time. For this reason, an efficient system is proposed in this paper to solve the problem of manual attendance. This system takes attendance electronically with the help of a fingerprint recognition system, and all the records are saved for subsequent operations. Staff biometric attendance system employs an automated system to calculate attendance of staff in an organization and do further calculations of monthly attendance summary in order to reduce human errors during calculations. In essence, the proposed system can be employed in curbing the problems of lateness, buddy punching and truancy in any institution, organization or establishment. The proposed system will also improve the productivity of any organization if properly implemented.
Local Binary Pattern and Ant Colony Optimization Based Feature Dimensionality Reduction Technique for Face Recognition Systems.
(Journal of Advances in Mathematics and Computer Science, 2015-04-13) Babatunde, R.S, Olabiyisi, S. O., Omidiora, E. O. and Ganiyu, R. A. (2015)
Feature dimensionality reduction is the process of minimizing the number of features in high dimensional feature space. It encompasses two vital approaches: feature extraction and feature selection. In face recognition domain, widely adopted face dimensionality reduction techniques include Principal component analysis, Discrete wavelet transform, Linear discriminant analysis and Gabor filters. However, the performances of these techniques are limited by strict requirement of frontal face view, sensitivity to signal shift and sample size, computational intensiveness amongst others. In this paper, a feature dimensionality reduction technique that employed Local binary pattern for feature extraction and Ant colony optimization algorithms for the selection of optimal feature subsets was developed. The developed technique identified and selected the salient feature subsets capable of generating accurate recognition. The average training time, recognition time and recognition rate obtained from the experiment on locally acquired face data using cross-validation evaluation approach indicate an efficient performance of the potential combination of both methods in a two-level technique for dimensionality reduction
: Assessing the performance of Random Partitioning and K-Fold Cross Validation methods of evaluation of a Face Recognition System.
(Journal of Advances in Image and Video Processing, 2015-06-08) Babatunde, R. S, Olabiyisi S. O, Omidiora E. O, Ganiyu, R. A. and Isiaka, R. M. (2015)
Face recognition has been an active research area in the pattern recognition and computer vision domains due to its many potential applications in surveillance, credit cards, passport and security. However, the problem of correct method of partitioning the face data into train and test set has always been a challenge to the development of a robust face recognition system. The performance of the System was tested on locally acquired face database when the face database was randomly partitioned and when k-fold Cross Validation partition was used. The face database was captured under the condition of significant variations of rotation, illumination and facial expression. Quantitative evaluation experimental results showed that Random Sampling technique has a higher average recognition rate (96.7%) than Cross Validation partition method (95.3%). However, recognition time in Cross Validation is faster (0.36 secs) than that of Random Sampling (0.38 secs).
Handwritten Character Recognition using Brainnet Library,
(Annals. Computer Science Series Journal, Published by Faculty of Computers and Applied Computer Science, Tibiscus University of Trinisoara, Romania., 2016-02-15) Babatunde, A. N., Abikoye, O.C., Babatunde, R.S. and Kawu, R.O. (2016)
Handwriting has continued to persist as a means of communication and recording information in dayto- day life even with the introduction of new technologies. Given its ubiquity in human transactions, machine recognition of handwriting has practical significance, as in reading handwritten notes in a PDA, in postal addresses on envelopes, in amounts in bank checks, in handwritten fields in forms, etc. An off-line handwritten alphabetical character recognition system using multilayer feed forward neural network is described, and a method, called, diagonal based feature extraction is used for extracting the features of the handwritten alphabets. This project implements this methodology using BrainNet Library. Ten data sets, each containing 26 alphabets written by various people, are used for training the neural network and 130 different handwritten alphabetical characters are used for testing. The proposed recognition system performs quite well yielding higher levels of recognition accuracy compared to the systems employing the conventional horizontal and vertical methods of feature extraction. This system, if modified will be suitable for converting handwritten documents into structural text form and recognizing handwritten names.
A novel approach to outliers removal in a noisy numeric dataset for efficient Mining.
(Published by Faculty of Computing and Information Systems, University of Ilorin., 2016-02-16) Ajiboye, A. R, Adewole, K. S., Babatunde, R. S. and Oladipo, I. D. (2016):
Data pre-processing is a key task in the data mining process. The task generally consumes the largest portion of the total data engineering effort while unveiling useful patterns from datasets. Basically, data mining is about fitting descriptive or predictive models from data. However, the presence of outlier sometimes reduces the reliability of the models created. It is, therefore, essential to have raw data properly pre-processed before exploring them for mining. In this paper, an algorithm that detects and removes outliers in a numeric dataset is proposed. In order to establish the effectiveness of the proposed algorithm, the clean data obtained through the implementation of the proposed approach is used to create a prediction model. Similarly, the clean data obtained through the use of one of the existing techniques is also used to create a prediction model. Each of the models created is simulated using a set of untrained data and the error associated with each model is measured. The resulting outputs from the two approaches reveal that, the prediction model created using the output from the proposed algorithm has an error of 0.38, while the prediction model created using the cleaned data from the clustering method gives an error of 0.61. Comparison of the errors associated with the models created using the two approaches shows that, the proposed algorithm is suitable for cleaning numeric dataset. The results of the experiment also unveils that, the proposed approach is efficient and can be used as an alternative technique to other existing cleaning methods.
Implications of System Identification Techniques on ANFIS E-learners Activities Models-A Comparative Study.
(Transactions on Machine Learning and Artificial Intelligence,, 2016-04-11) Isiaka, R. M., Omidiora, E. O. Olabiyisi, S. O., Okediran O.O. and Babatunde, R. S. (2016):
Efficient e-learners activities model is essential for real time identifications and adaptive responses. Determining the most effective Neuro- Fuzzy model amidst plethora of techniques for structure and parameter identifications is a challenge. This paper illustrates the implication of system identification techniques on the performance of Adaptive Network based Fuzzy Inference System (ANFIS) E-learners Activities models. Expert knowledge and Historical data were used to formulate the system and their performances were compared. Similarly, comparison was made between memberships functions selected for Historical data identification. The efficiencies of the simulated models in MATLAB editor were determined using both classification uncertainty metrics and confusion matrix–based metrics. The classification uncertainty metrics considered are Mean Absolute Error (MAE) and Root Mean-Squared Error (RMSE). The confusion matrix-based metrics used are Accuracy, Precision and Recall. It was discovered that the model based on Experts Knowledge after training outperformed those based on Historical Data. The performances of the Membership Functions after ranking are Sigmoid, Gaussian, Triangular and G-Bells respectively.
Gender Recognition using Local Binary Pattern and Naïve Bayes Classifier
(African Journal Online. Published by Nigeria Computer Society Journal. Journal of Computer Science and its Applications, 2016-10-24) Babatunde, R. S., Abdulsalam, S. O., Yusuff, S. R., and Babatunde, A. N. (2016):
Human face provides important visual information for gender perception. Ability to recognize a particular gender is very important for the purpose of differentiation. Automatic gender classification has many important applications, for example, intelligent user interface, surveillance, identity authentication, access control and human-computer interaction amongst others. Gender recognition is a fundamental task for human beings, as many social functions critically depend on the correct gender perception. Consequently, real-world applications require gender classification on real-life faces, which is much more challenging due to significant appearance variations in unconstrained scenarios. In this study, Local Binary Pattern is used to detect the occurrence of a face in a given image by reading the texture change within the regions of the image, while Naive Bayes Classifier was used for the gender classification. From the results obtained, the gender correlation was 100% and the highest accuracy of the result obtained was 99%.The system can be employed for use in scenarios where real time gender recognition is required.
An Evaluation of performance of k-NN and Mahalanobis Distance on Local Binary Pattern Feature representation for Access Control in Face Recognition System.
(Computing, Information Systems & Development Informatics Journal, 2017-04-11) Yusuff, S.R., Balogun, F.B., Babatunde, A.N. and Babatunde, R.S. (2017)
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 k- NN with 94.9% and (0.392 and 2.895) respectively. This result demonstrates the robustness of Mahalanobis distance classifier for use access control system.
Implication of variation in Ant Colony Optimization Algorithm (ACO) parameters on Feature Subset Size.
(Computing and Information System Journal, University of West Scotland, 2017-05-20) Babatunde, R. S., Isiaka, R.M., Adigun, O.J. and Babatunde, A.N. (2017)
Purpose: The purpose of this research is to apply ant colony optimization (ACO), a nature inspired computational (NIC) technique to achieve optimal feature subset selection, for the purpose of data dimensionality reduction, which will remove redundancy, provide a reduced storage space, improve memory utilization and subsequently, classification task. Design/Methodology/Approach: The ACO feature subset selection process was filter-based, accomplished by using correlation coefficient as the heuristic information 􀟟 which guides the search to determine the pixel (feature) attractiveness. The intensity of the pixel represents the pheromone level 􀟬. Also, ρ, φ, 􀟙 and 􀟚are key parameters which were tuned to various arbitrary values at each iteration to determine the best combination of parameters which can result in a reduced feature subset (window size). The procedure was simulated in MATLAB 2012. Findings: The study has been able to demonstrate that experimental results obtained by tuning the parameters of ACO at each run gave a reduced feature subset (window size), at parameter settings of α (0.10), β (0.45), ρ (0.50), φ (1.0) which was salient for efficient face recognition thereby demonstrating the effectiveness and superiority of filter-based over wrapper–based feature selection. Research Limitation: The major limiting factor in this research was time and materials which hindered the further testing and implementation of innovative computational intelligence. Originality/Value: This research work is valuable because it gives the implication of the variation of the ACO parameters which determines the extent to which the optimal solution construction that results in a reduced feature subset can be realized.
A Systematic Review of Feature Dimensionality Reduction Techniques for Face Recognition Systems.
(Journal of Digital Innovations & Contemp Res. In Sc., Eng & Tech., 2017-09-07) Babatunde, R.S. (2017)
In this research, a comparison of the various techniques used for face recognition is shown in tabular format to give a precise overview of what different authors have already projected in this particular field. A systematic review of 40 journal articles pertaining to feature dimensionality reduction was carried out. The articles were reviewed to appraise the methodology and to identify the key parameters that were used for testing and evaluation. The dates of publication of the articles were between 2007 and 2015. Ten percent (10%) of the articles reported the training time for their system while twenty four percent (24%) reported their testing time. Sixty seven percent (67%) of the reviewed articles reported the image dimension used in the research. Also, only forty eight percent (48%) of the reviewed articles compared their result with other existing methods. The main emphasis of this survey is to identify the major trade-offs of parameters and (or) metrics for evaluating the performance of the techniques employed in dimensionality reduction by existing face recognition systems. Findings from the review carried out showed that major performance metrics reported by vast amount of researchers in this review is recognition accuracy in which eighty six percent (86%) of the authors reported in their experiment.
A Generalized Neuron Model (GNM) Based Human Age Estimation.
(Advances in Multidisciplinary & Scientific Research Journal., 2017-10-20) Babatunde, R.S. and Yakubu, I.A. (2017)
In real-world application, the identification characteristic of face images has been widely explored, ranging from National ID card, International passport, driving license amongst others. In spite of the numerous investigation of person identification from face images, there exists only a limited amount of research on detecting and estimating the demographic information contained in face images such as age, gender, and ethnicity. This research aim at detecting the age/age range of individual based on the facial image. In this research, a generalized neuron (GN), which is a modification of the simple neuron, is used, to overcome some of the problems of artificial neural network (ANN) and improve its training and testing performance. The GN is trained with discrete wavelet transform (DWT) features obtained after the application of Canny edge detection algorithm on the face Image. Validating the technique on FGNET face images reveals that the frequency domain features obtained using the DWT captures the wrinkles on the face region, which represents a distinguishing factor on the face as humans grow older. The empirical results demonstrates that the GN outperforms the simple neuron, with detection rate of 93.5%, training time of 96.30secs, matching time of 14secs and root mean square error of 0.0523. The experimental results suggest that the GN model performs comparably and could be adopted for detecting human ages.
Image Encryption System based on Length Three Moduli Set
(Foundation of Computer Science (FCS), NY, USA, 2018-05-01) Damilare Peter Oyinloye
Digital images have found usage in almost all our everyday applications. These images sometimes contain confidential and intelligible information which need to be protected when stored on memory or transmitted over networks (Intranet or Internet). Many techniques have been proposed to deal with this security issues in the past. This paper proposes a simple scrambling algorithm to encrypt and decrypt the grey level image based on random number generation and Residue Number System (Forward and Reverse Conversion). The image is first encrypted by changing the position of each pixel in the original image without changing the value of grey level. The original image reads row by row, pixel by pixel and each pixel will take a new position in the scrambled image. The new position is chosen based on random number generation from the random number generator. The key will be generated as a matrix during the encryption process and also the key saves the position of each pixel in the encrypted/scrambled image. The encryption layer transforms the scrambled image to moduli images which automatically adds an extra security layer to our data. The encrypted moduli images is decrypted by decoding the moduli images and converting them back to a single scrambled image (Reverse Conversion) and the single scrambled image back to the Plain and Original Image by using the saved key matrix. This scheme achieves an enhanced image encryption process and a more efficient decryption process without loses of any inherent information of the recovered plain image.
An Improved Image Scrambling Algorithm Using {2 n -1, 2 n , 2 n +1}
(University of the West of Scotland, School of Computing, 2018-10-01) Damilare Peter Oyinloye; Kazeem Alagbe Gbolagade
Traditional iris segmentation methods and strategies regularly contain an exhaustive search of a large parameter space, which is sensitive to noise, time-consuming and no longer secured enough. To address these challenges, this paper proposes a secured iris template. This paper proposes a technique to secure the iris template using steganography. The experimental analysis was carried out on matrix laboratory (MATLABR2015A) environment. The segmented iris region was normalized to decrease the dimensional inconsistencies between iris region areas with the aid of the usage of Hough transform (HT). The features of the iris were encoded by convolving the normalized iris region with 1D Log- Gabor filters in order to generate a bit-wise biometric template. Then, least significant bit (LSB) was used to secure the iris template. The Hamming distance was chosen as a matching metric, which gives the measure of how many bits disagreed between the templates of the iris. The system operated at a very good training time and high level of conditional testing signifying high optimization with recognition accuracy of 92% and error of 1.7%. The proposed system is reliable, secure and efficient with the computational complexity significantly reduced. The proposed technique provides an efficient approach for securing the iris template.
Comparison Performance Evaluation of Modified Genetic Algorithm and Modified Counter Propagation Network for Online Character recognition.
(Transactions on Machine Learning and Artificial Intelligence., 2018-10-07) Adigun O. J., Fenwa, O. D. and Babatunde, R. S. (2018)
This paper carries out performance evaluation of a Modified Genetic Algorithm (MGA) and Modified Counter Propagation Network (MCPN) Network for online character recognition. Two techniques were used to train the feature vectors using supervised and unsupervised methods. First, the modified genetic algorithm used feature selection to filter irrelevant features vectors and improve character recognition because of its stochastic nature. Second, MCPN has its capability to extract statistical properties of the input data. MGA and MCPN techniques were used as classifiers for online character recognition then evaluated using 6200-character images for training and testing with best selected similarity threshold value. The experimental results of evaluation showed that, at 5 x 7 pixels, MGA had 97.89% recognition accuracy with training time of 61.20ms while MCPN gave 97.44% recognition accuracy in a time of 62.46ms achieved. At 2480, MGA had 96.67% with a training time of 4.53ms, whereas MCPN had 96.33% accuracy with a time of 4.98ms achieved. Furthermore, at 1240 database sizes, MGA has 96.44 % recognition accuracy with 0.62ms training time whereas MCPN gave 96.11% accuracy with 0.75ms training time. The two techniques were evaluated using different performance metrics. The results suggested the superior nature of the MGA technique in terms of epoch, recognition accuracy, convergence time, training time and sensitivity.
Adaptive Neuro Fuzzy Inference System (ANFIS) Based Detection of Cervical Cancer.
(Published by LAUTECH, Ogbomoso. Oyo State, 2018-11-22) Babatunde, R. S. and Muhammad-Thani S. (2018):
Cancer occurs when cells in an area of the body grows abnormally. Cervical cancer is a cancer that is known to be the second most deadly cancer affecting women worldwide which emerges from the cervix of a woman. The normal cervix has two main types of cells: squamous cells, that protect the outside of the cervix and glandular cells. Cervical cancer is caused by abnormal changes in either of these cell types in the cervix. Pap smear screening is the most successful attempt of medical science and practice for the early detection of cervical cancer. However, the manual analysis of the cervical cells (pap smear) is time consuming, laborious and error prone. In this research, detection of cervical cancer was done using Adaptive Neuro-Fuzzy Inference System (ANFIS), a hybrid intelligent system which combines the fuzzy logic qualitative approach and adaptive neural network capabilities. The system was modelled to predict the probability of cervical cancer as either high probability or low probability based on the input risk factors of cervical cancer as contained in the dataset used. The efficiency of the ANFIS in predicting cervical cancer was tested using a dataset collected from Leah Foundation, a non-governmental organization in Nigeria, having a total of 250 patient’s data. The dataset contains eight risk factors of cervical cancer, which serves as the input variables. The output is the classified risk factors each patient is prone to. A patient with low level (represented as 1) is classified to be of no or less probability of having cervical cancer, while a patient with high level (represented as 2) is classified to be a victim of high probability of having cervical cancer. The result of this system shows that the system can be adopted by medical practitioners as well as for individual use.
A Comparative analysis of Texture feature and 3D colour Histogram for Content Based Image Retrieval.
(U6+ Consortium of African Universities Proceedings. Maiden Edition on Harnessing African Potentials for Sustainable Development, Calabar, Nigeria., 2019-09-15) Babatunde, R. S., and Ajao, J. F. (2019)
The visual content of an image can be used to search for similar images based on user interest from a large database, in the art known as Content Based Image Retrieval (CBIR). In CBIR systems, the user presents a query image to the system, the system obtains the feature vector, which is compared with certain image features of the images in the database. The adequacy of the feature vectors extracted for the retrieval of appropriate and exact image from the database is an open issue which calls for continual research and attention. This work involved the retrieval of similar images based on the content of the image by comparing the feature vectors of the queried image and those of the image database using Mahalanobis distance measure. Textural feature vectors of the images were obtained using local binary pattern and 3D colour histogram feature vectors were also extracted. The performances of these two different feature vectors were compared based on the distance metrics to determine their suitability. The similar images in the database are displayed along with their similarity distance value, in which the minimum distance is a metric for the matched images. The CBIR system was implemented using a locally acquired database of over 600 different images. Various images were used to query the CBIR system, which was able to successfully output similar images. The simulation result obtained revealed that textural feature vectors are more adequate in terms of speed and accuracy in content-based image retrieval than 3D colour histogram based on the images used in the experiment.
Performance Analysis of Gray Code Number System in Image Security
(2019-10-03) Babatunde, A.N., Jimoh, E.R, Oshodi, O. & Alabi, O.A.
The encryption of digital images has become essential since it is vulnerable to interception while being transmitted or stored. A new image encryption algorithm to address the security challenges of traditional image encryption algorithms is presented in this research. The proposed scheme transforms the pixel information of an original image by taking into consideration the pixel location such that two neighboring pixels are processed via two separate algorithms. The proposed scheme utilized the Gray code number system. The experimental results and comparison shows the encrypted images were different from the original images. Also, pixel histogram revealed that the distribution of the plain images and their decrypted images have the same pixel histogram distributions, which means that there is a high correlation between the original images and decrypted images. The scheme also offers strong resistance to statistical attacks. Keywords: Gray Code, Number System, Spatial domain, Image encryption.
A Machine Learning Approach to Dropout Early Warning System Modeling. International
(Journal of Advanced Studies In Computer Science And Engineering, 2019-10-18) Isiaka R.M., Babatunde R. S., Ajao F.J. and Abdulsalam S.O. (2019)
Dropout has been identified as a social problem with an immediate and long term consequences on the socioeconomic development of the society. A major element of a conceptual framework for dropout monitoring and control system is the prevention component. The component specified the need for automatic dropout early warning system. This paper presents the procedure for building the adaptive model, that is the core element for the realization of the prevention component of the framework. The model development is guided by the knowledge of the domain experts. An experimental approach was used to identify the K-Nearest Neighbors (KNN) as the best of the six algorithms considered for the adaptive models. Other algorithms explored are Logistic Regression (LR), Linear Discriminant Analysis (LDA), Classification and Regression Trees (CART), Gaussian Naive Bayes (NB) and Support Vector Machines (SVM). Historical dataset from a State University in the North Central region of Nigeria was used for building and testing the adaptive model. The average performance of the model on full attributes and the first year academic attributes compared favourably with 0.98 Precision, Recall and F1-Scores. Conversely, the Precision, Recall and F1-Scores on the entry attributes are 0.90, 0.95 and 0.92 respectively. The implication of this is that the academic performance is a very significant factor for dropping out of the educational system. The developed model can provide an early signal on the students with the propensity to dropout, thereby serving as an advisory system for the students, parents and school management towards curtailing the menace.
Comparative Approach of Back Propagation Neural Network and Decision Tree on Breast Cancer Classification. .
(Proceedings of the 8th International Conference on Mobile e-Services (ICoMeS) – Lautech, Ogbomoso. Oyo State. Nigeria, 2019-10-18) Babatunde, R. S, Adewole, K. S., and Ajiboye, A. R. (2019)
The use of data mining methods in incorporating decision making has been increasing in the past decades. Data mining simply refers to extracting or mining knowledge from large amount of data. Over the years, medical image processing has benefited immensely from data mining techniques including breast cancer diagnosis. Sonography (also known as ultrasound) has become a great addition to mammography and magnetic resonance imaging (MRI) as imaging techniques dedicated to providing breast cancer screening. This technique is time-consuming and often characterized with low accuracy. Hence, the need to develop a robust classification model with high performance accuracy and reduced false alarm. In this paper, the performance of back propagation neural networks (BPNN) and C4.5 decision tree (DT) for breast cancer prediction was carried out. Filter based feature selection approach using correlation filter was employed for ranking features according to their predictive power. The model was simulated using WEKA data mining tools and extensive comparative study was performed based on the standard evaluation metrics. The performance of the two classifiers was compared based on their predictive accuracy, precision, recall, kappa statistic and other relevant statistical measures. The simulation results shows that C4.5 outperforms BPNN in terms of training time (0.16 secs) and accuracy (94.2857% ) while BPNN has 46.9secs training time and accuracy of 90.9524%. However, the result also reveals that BPNN outperforms C4.5 in terms of error rate, with BPNN having mean absolute error of 0.0542 while C4.5 has mean absolute error of 0.0834. It can therefore be deduced from the comparison that C4.5 can be a good option for prediction task considering the fast training time of the algorithm as well as the high accuracy of prediction.

Browse

Browsing Depatment of Computer Science by Issue Date

Results Per Page

Sort Options