ICTACT Journal on Soft Computing

Sunday Olusanya Olatunji ¹, Hossain Arif ²

Affiliations
1 Department of Computer Science, Adekunle Ajasin University, NG
2 School of Engineering and Computer Science, BRAC University, BD

Abstract

In this work, a new identification model, based on extreme learning machine (ELM), to better identify Erythemato - Squamous skin diseases have been proposed and implemented and the results compared to that of the classical artificial neural network (ANN). ELMs provide solutions to single- and multi- hidden layer feed-forward neural networks. ELMs can achieve high learning speed, good generalization performance, and ease of implementation. Experimental results indicated that ELM outperformed the classical ANN in all fronts both for the training and testing cases. The effect of varying size of training and testing set on the performance of classifiers were also investigated in this study. The proposed classifier demonstrated to be a viable tool in this germane field of medical diagnosis as indicated by its high accuracy and consistency of result.

Keywords

Extreme Learning Machine, Artificial Neural Network, Erythemato-Squamous Skin Diseases.

Full Text

     

Sunday Olusanya Olatunji ¹

Affiliations
1 Department of Computer Science, Adekunle Ajasin University, Akungba Akoko, Ondo State, NG

Abstract

This paper presented two maintainability prediction models that are developed and compared for object-oriented software systems based on the recently introduced learning algorithm called Sensitivity Based Linear Learning Method (SBLLM) and extreme learning machines (ELM). As the number of object-oriented software systems increases, it becomes more important for organizations to maintain those systems effectively. However, currently only a small number of maintainability prediction models are available for object oriented systems. The model was constructed using popular object-oriented metric datasets, collected from different object-oriented systems. Prediction accuracy of the models were evaluated and compared with each other and with other commonly used regression-based models and also with Bayesian network based model which were earlier developed using the same datasets. Empirical results from simulation show that the proposed ELM and SBLLM based models produced promising results in term of prediction accuracy measures authorized in OO software maintainability literatures, better than most of the other earlier implemented models on the same datasets.

Keywords

Sensitivity Based Linear Learning Method (SBLLM), Extreme Learning Machines, Object Oriented Software Systems, Software Metrics, Software Maintainability Prediction Models.

Full Text