China
Africa
Explore chapters and articles related to this topic
Modelling and analysis of skin pigmentation
Published in Ahmad Fadzil Mohamad Hani, Dileep Kumar, Optical Imaging for Biomedical and Clinical Applications, 2017
Ahmad Fadzil Mohamad Hani, Hermawan Nugroho, Norashikin Shamsudin, Suraiya H. Hussein
The drawbacks of the crossing are related to that the product/children of one or two individuals with weak fitness is/are not alike. To avoid the effect of these numerous crossings, the operation can be directed towards the selection of individuals with good fitness to be parents. This process not only tends to limit the exploration of the search space but also may lead to premature convergence. The crossing of parents with close genes information rarely provides new information in the population, affecting the algorithm convergence speed. The crossing probability of an individual ranges from 0.6 to 0.95 [178,179]. Figure 4.52 shows the illustration of crossing operation between two parents.
A new variant of deep belief network assisted with optimal feature selection for heart disease diagnosis using IoT wearable medical devices
Published in Computer Methods in Biomechanics and Biomedical Engineering, 2022
Fathima Aliyar Vellameeran, Thomas Brindha
A novel hybrid PS-GWO algorithm is developed to optimize the features and for the optimization of ‘number of hidden neurons and activation function’ of DBN to attain better diagnosis results and accuracy. This PS-GWO algorithm is proposed by adopting the features of both approaches like GWO (Mirjalili et al. 2014) and PSO (Pedersen and Chipperfield 2010) algorithms. This new PS-GWO algorithm updates the positions by considering both PSO and GWO algorithms. GWO has several features like simplicity, use of less control parameters, and simpler implementation and ability of solving the diverse optimization problems. However, this approach suffers from premature convergence, lack of diversity in population and the imbalance among the exploration and exploitation phases. Thus, a PSO algorithm is used for solving the problems of GWO algorithm. PSO has features like computational efficiency, easy and simpler implementation, robustness toward control parameters and offers faster convergence. Thus, the proposed PS-GWO algorithm updates the position of alpha
Directed jaya algorithm for delivering nano-robots to cancer area
Published in Computer Methods in Biomechanics and Biomedical Engineering, 2020
Doaa Ezzat, Safaa Amin, Howida A. Shedeed, Mohamed F. Tolba
Despite the efficiency of these algorithms, many drawbacks appeared when using these algorithms in solving optimization problems. Some of these drawbacks are as follows (Selvi and Umarani 2010; Odili et al. 2018):Particle Swarm Optimization (PSO) may be trapped in a local optimum and can't reach the global optimum.Time for convergence in Ant Colony Optimization (ACO) is uncertain.Genetic Algorithms (GAs) have a lot of limitations. For some problems, GAs may require several hours to several days for convergence. GA’s do not scale well with complexity. In many problems, GA’s may converge to local optima or even arbitrary points rather than the global optimum.Cuckoo search (CS) uses several parameters and this affects its speed, and it sometimes falls into local optima.The speed of Bat Algorithm (BA) is restricted by several parameters.In complex problems, Teaching-Learning-Based Optimization (TLBO) has been found to perform very poorly and this leads to premature convergence.
Parametrizing the Kepler exoplanet period-radius distribution with the bivariate normal inverse Gaussian distribution
Published in Journal of Applied Statistics, 2019
At the beginning of the PSO procedure, the positions of the particles of the swarm D is the search region, and D, it is pulled back to the boundary of D. Moreover, the limiting velocity of the particle acts to prevent the PSO from getting stuck in a local maximum, or its premature convergence. For each PSO iteration, the objective function 4), is calculated to determine the personal best position ω refers to the inertia weight which modulates the influence of the former velocity on the current velocity. The terms 23], and