The aim of this study is to evaluate the utility an artificial neural network (ANN) model in diagnosing idiopathic generalized epilepsy (IGE) and compare results diagnostic constructed by combining expression levels miR‐146a, miR‐155, miR‐132 genes using ANN, random forest (RF), discriminant analysis (DA). qRT‐PCR employed determine three miRNA genes. Forty‐six IGE patients 51 healthy controls were included study. Three genetic biomarkers assess discriminative power disease, they combined...

Feed-forward neural networks have been frequently used in forecasting problems, recently. In this study, we propose a naive method to improve the ability of feed-forward with single hidden layer by adapting meta fuzzy functions. Because are very sensitive initial random weights, usually some numbers repeats processed different weights. The forecasts for are, then, averaged equal weights obtain more reliable results. However, if can assign correct initials appropriate produce competitive...

Abstract In order to analyse the currency crises in Turkey over period of January 1990 and October 2019, we first dated with meta‐possibilistic fuzzy index functions. Then, determined significant predictors or leading indicators crisis logistic regression. Finally, tried measure compare in‐sample out‐of‐sample performances our method an generated by principal component analysis (PCA). We found that models using have higher than PCA. concluded change real exchange rate, bank loans deposits,...

Meta-analysis was introduced to aggregate the findings of different primary studies in statistical aspects. However, proposed study, term "meta" is used models for a specific topic with help fuzzy c-means clustering method. One motivations method based on concept indices. In literature, there are numerous indices under conditions purpose. Our assumption that each index has some information given dataset. Therefore, meta functions, which include function certain degree membership value,...

Intuitionistic meta fuzzy forecast combination functions are introduced in the paper. There two challenges literature, determining optimum weights and methods to combine. Although there a few studies on methods, numerous of forecasting methods. In this sense, questions like “What should we choose combination?” function or for methods” handled proposed method. Thus, first contributions that paper aims propose obtain proper by employing (MFFs). MFFs recently aggregating different specific...

Time series models are used extensively in many fields, such as medicine, engineering, business, economics, and finance, with the aim of making forecasts through help observation values from previous periods.Therefore, there efforts to improve time forecasting performances recent literature, mainly using alternative/non-probabilistic methods.In present study, a novel approach has been proposed by combining type-1 fuzzy functions (T1FF) Autoregressive moving average (ARMA) model based on grey...