A survey on computational intelligence approaches for predictive modeling in prostate cancer

Abstract

Predictive modeling in medicine involves the development of computational models which are capable of analysing large amounts of data in order to predict healthcare outcomes for individual patients. Computational intelligence approaches are suitable when the data to be modelled are too complex for conventional statistical techniques to process quickly and efficiently. These advanced approaches are based on mathematical models that have been especially developed for dealing with the uncertainty and imprecision which is typically found in clinical and biological datasets. This paper provides a survey of recent work on computational intelligence approaches that have been applied to prostate cancer predictive modeling, and considers the challenges which need to be addressed. In particular, the paper considers a broad definition of computational intelligence which includes evolutionary algorithms (also known as ing the performance of prostate cancer predictive models, and the suitability of these approaches are discussed.

An extensive evaluation of seven machine learning methods for rainfall prediction in weather derivatives

“…The application of seven machine learning algorithms reported in this work shows that intelligent systems are useful complementing the results reported by Liao et al (2012); Patel et al (2015); Cosma et al (2017), who all show the benefit that machine learning-based intelligent systems can have on real-life scenarios.…”

“…Moreover, we can apply a range of machine learning (data mining) techniques to solve real-life problems (Liao et al, 2012;Patel et al, 2015;Cosma et al, 2017). The process of predicting rainfall and pricing falls under this concept, where a machine learning based intelligent system can either discover a set of rules or provide an equation.…”

Deep stacked sparse auto-encoders for prediction of post-operative survival expectancy in thoracic lung cancer surgery

“…where: n = the number of data inputs λ = the weight degradation parameter which controls the relative momentousness of the prime and the second terms β = a parameter controlling the weight of the sparse penalty term m = the number of neurons in the hidden layer KL is the kullback-leibler (Cosma et al, 2017) divergence between tow bernoulli random variables with mean ρ and mean is the average activation of hidden unit j or averaged over the training set and ρ is a sparsity parameter, usually a small value near to zero (ρ = 0.05)…”

“…Past research evidences that soft computing techniques such as artificial neural networks (Geng et al, 2016;Luo, 2017), fuzzy logic (Kuo et al, 2015;Omiotek et al, 2013), support vector machines (Dobrowolski et al, 2016;Dolatabadi et al, 2017) and deep learning (Ciompi et al, 2015;Cosma et al, 2017;Gargeya and Leng, 2017;Pham et al, 2017;Sun et al, 2017;Yin et al, 2016;Zhang et al, 2016a), which have been used for medical prediction and decision-making, have good results using training data sets. Simonsen et al (2015) have expressed risk factors for post-operative pneumonia (POP) in patients undergoing therapy by lung cancer surgery within 30 days after surgery, with a logistic regression method.…”

Optimization of Efficient Route and Profitable Fish Aggregating Device Based on Firefly and Genetic Algorithms in Maritime

“…In this paper, two optimization methods (FA and GA) already been applied to various problems were evaluated. GA was selected because it was suitable for handling discrete problems, in which the optimum solution was determined from a finite number of possible solutions [16]. GA exhibited reliable problem solving in many applications, and could improve the significant parameter for forecasting tools, face recognition, and simulation of some diseases [17][18][19].…”