Hybrid journal (It can contain Open Access articles) ISSN (Print) 1751-648X - ISSN (Online) 1751-6498 Published by Inderscience Publishers[449 journals]

Authors:Sahar Ardalan, Sama Goliaei, Ayaz Isazadeh Pages: 1 - 8 Abstract: Signal machine is an abstract geometrical model of computation, which can be viewed as a continuous space and time generalisation of cellular automata. Almost all studies that have been made are about deterministic signal machines. In spite of few studies that have been made on non-deterministic signal machines, the present paper shows their high efficiency in solving problems using a well-known combinatorial problem. We provide a method to solve the graph dominating set problem using non-deterministic signal machines. First, we show how to design a signal machine for each specific instance of the dominating set problem. Then, we propose a signal machine which solves the dominating set problem for any instance of the problem, and show how to reduce the space complexity of solution using non-determinism. Keywords: abstract geometrical computation; non-deterministic signal machines; dominating set problem; computational model Citation: International Journal of Innovative Computing and Applications, Vol. 11, No. 1 (2020) pp. 1 - 8 PubDate: 2020-02-24T23:20:50-05:00 DOI: 10.1504/IJICA.2020.105311 Issue No:Vol. 11, No. 1 (2020)

Authors:Ankit Thakkar, Dhara Mungra, Anjali Agrawal Pages: 9 - 29 Abstract: The proliferated increase in the commercial benefits of sentiment analysis accumulated a huge interest in the domain of sentiment classification. Sentiment analysis categorises a given text into positive or negative class. In this paper, we present an empirical comparison between different training algorithms gradient descent (GD), gradient descent with momentum backpropagation (GDM), gradient descent adaptive learning rate backpropagation (GDA), gradient descent with momentum and adaptive learning rate backpropagation (GDX), and Levenberg-Marquardt backpropagation (LM), used for training the neural network for the domain of sentiment classification. The performance of all the methods is compared and evaluated using three balanced binary datasets from various domains with different features using various performance metrics such as accuracy, precision, recall, <i>f</i>-score, mean squared error, and training time. The experiments are performed five times with different random seed values using 10-fold cross-validation. The results indicate that GDX and LM outperform other methods in terms of classification accuracy. Keywords: sentiment analysis; artificial neural network; training algorithms; binary class; different domains Citation: International Journal of Innovative Computing and Applications, Vol. 11, No. 1 (2020) pp. 9 - 29 PubDate: 2020-02-24T23:20:50-05:00 DOI: 10.1504/IJICA.2020.105315 Issue No:Vol. 11, No. 1 (2020)

Authors:Anthony Brabazon, SeÃ¡n McGarraghy Pages: 30 - 45 Abstract: The metaphor of 'foraging as search' provides a rich source of inspiration for the design of optimisation algorithms. An extensive literature has resulted in computer science over the past twenty years based on this, with algorithmic families such as ant colony optimisation and honeybee optimisation amongst others, being successfully applied to a wide range of real-world problems. Of course, all organisms must forage to acquire necessary resources and in recent years, increasing attention has been paid to the mechanisms by which non-neuronal organisms, in other words organisms without a central nervous system, forage. The vast majority of living organisms, encompassing some 99.5% of all biomass on earth, are non-neuronal. In this paper we introduce the plasmodial slime mould <i>Physarum polycephalum</i>. This non-neuronal organism is formed when individual amoebae swarm together and fuse, resulting in a large bag of cytoplasm encased within a thin membrane which acts a single organism. Inspiration has been drawn from some of its foraging behaviours to develop algorithms for graph optimisation and exemplars of these algorithms along with some suggestions for future research are presented in this paper. Keywords: slime mould algorithms; foraging-inspired algorithms; graph optimisation; non-neuronal organisms Citation: International Journal of Innovative Computing and Applications, Vol. 11, No. 1 (2020) pp. 30 - 45 PubDate: 2020-02-24T23:20:50-05:00 DOI: 10.1504/IJICA.2020.105316 Issue No:Vol. 11, No. 1 (2020)

Authors:Anthony Brabazon, SeÃ¡n McGarraghy Pages: 46 - 60 Abstract: This paper investigates the classical generalised assignment problem (GAP), a challenging combinatorial optimisation problem that arises in numerous applications and that has attracted a great deal of research. For solving it we propose a hybrid metaheuristic combining guided search (GS), iterated local search (ILS), and very large-scale neighbourhood search (VLSN). The hybrid method is iterative. It starts with a random assignment, and in every iteration it acts in the following way: 1) The best current assignment is perturbed. 2) An exponential size neighbourhood of the perturbed assignment is constructed. It is the feasible solution set of a special GAP where only two fixed machines can execute a job. The neighbourhood construction is guided by a technique penalising poor machine/job selections. 3) The exponential neighbourhood is searched for improvement. Exploring the neighbourhood amounts to solving a monotone binary program (BP) – a monotone BP is one with two non-zero coefficients of opposite sign per column. We prove that the proposed metaheuristic runs in polynomial-time when applied to a variation of GAP. Good computational results in terms of solution quality, as well as of computation speed, are obtained with two new best values on hard instances. Keywords: generalised assignment problem; hybrid metaheuristic; very large scale neighbourhood; iterated local search; guided search; variable-fixing Citation: International Journal of Innovative Computing and Applications, Vol. 11, No. 1 (2020) pp. 46 - 60 PubDate: 2020-02-24T23:20:50-05:00 DOI: 10.1504/IJICA.2020.105317 Issue No:Vol. 11, No. 1 (2020)