Derya Avci

An Adaptive Network-Based Fuzzy Inference System (ANFIS) for the prediction of stock market return: The case of the Istanbul Stock Exchange

Stock market prediction is important and of great interest because successful prediction of stock prices may promise attractive benefits. These tasks are highly complicated and very difficult. In this paper, we investigate the predictability of stock market return with Adaptive Network-Based Fuzzy Inference System (ANFIS). The objective of this study is to determine whether an ANFIS algorithm is capable of accurately predicting stock market return. We attempt to model and predict the return on stock price index of the Istanbul Stock Exchange (ISE) with ANFIS. We use six macroeconomic variables and three indices as input variables. The experimental results reveal that the model successfully forecasts the monthly return of ISE National 100 Index with an accuracy rate of 98.3%. ANFIS provides a promising alternative for stock market prediction. ANFIS can be a useful tool for economists and practitioners dealing with the forecasting of the stock price index return.

Automatic hepatitis diagnosis system based on Linear Discriminant Analysis and Adaptive Network based on Fuzzy Inference System

In this paper, an automatic diagnosis system based on Linear Discriminant Analysis (LDA) and Adaptive Network based on Fuzzy Inference System (ANFIS) for hepatitis diseases is introduced. This automatic diagnosis system deals with the combination of feature extraction and classification. This automatic hepatitis diagnosis system has two stages, which feature extraction - reduction and classification stages. In the feature extraction - reduction stage, the hepatitis features were obtained from UCI Repository of Machine Learning Databases. Then, the number of these features was reduced to 8 from 19 by using Linear Discriminant Analysis (LDA). In the classification stage, these reduced features are given to inputs ANFIS classifier. The correct diagnosis performance of the LDA-ANFIS automatic diagnosis system for hepatitis disease is estimated by using classification accuracy, sensitivity and specificity analysis, respectively. The classification accuracy of this LDA-ANFIS automatic diagnosis system for the diagnosis of hepatitis disease was obtained in about 94.16%.

An expert system based on Generalized Discriminant Analysis and Wavelet Support Vector Machine for diagnosis of thyroid diseases

Nowadays, there are many persons, which suffer from thyroid diseases. Therefore, the correct diagnosis of these diseases are very important topic. In this study, a Generalized Discriminant Analysis and Wavelet Support Vector Machine System (GDA_WSVM) method for diagnosis of thyroid diseases is presented. This proposed system includes three phases. These are feature extraction - feature reduction phase, classification phase, and test of GDA_WSVM for correct diagnosis of thyroid diseases phase, respectively. The correct diagnosis performance of this GDA_WSVM expert system for diagnosis of thyroid diseases is estimated by using classification accuracy and confusion matrix methods, respectively. The classification accuracy of this expert system for diagnosis of thyroid diseases was obtained about 91.86%.

An intelligent diagnosis system for diabetes on Linear Discriminant Analysis and Adaptive Network Based Fuzzy Inference System: LDA-ANFIS

In this study, an intelligent diagnosis system for diabetes on Linear Discriminant Analysis (LDA) and Adaptive Network Based Fuzzy Inference System (ANFIS): LDA-ANFIS is presented. The structure of this LDA-ANFIS intelligent system for diagnosis of diabetes is composed by two phases: The Linear Discriminant Analysis (LDA) phase and classificiation by using ANFIS classifier phase. In first phase, Linear Discriminant Analysis (LDA) is used to separate features variables between healthy and patient (diabetes) data. In second phase, the healthy and patient (diabetes) features obtained in first phase are given to inputs of ANFIS classifier. The correct diagnosis performance of the LDA-ANFIS intelligent system is calculated by using sensitivity and specificity analysis, classification accuracy and confusion matrix respectively. The classification accuracy of this LDA-ANFIS intelligent system was obtained about 84.61%.

An expert system for speaker identification using adaptive wavelet sure entropy

In this study, an expert speaker identification system is presented for speaker identification using Turkish speech signals. Here, a discrete wavelet adaptive network based fuzzy inference system (DWANFIS) model is used for this aim. This model consists of two layers: discrete wavelet and adaptive network based fuzzy inference system. The discrete wavelet layer is used for adaptive feature extraction in the time-frequency domain and is composed of discrete wavelet decomposition and discrete wavelet entropy. The performance of the used system is evaluated by using repeated speech signals. These test results show the effectiveness of the developed intelligent system presented in this paper. The rate of correct classification is about 90.55% for the sample speakers.

An expert diagnosis system for classification of human parasite eggs based on multi-class SVM

In this paper, it is proposed a new methodology based on invariant moments and multi-class support vector machine (MCSVM) for classification of human parasite eggs in microscopic images. The MCSVM is one of the most used classifiers but it has not used for classification of human parasite eggs to date. This method composes four stages. These are pre-processing stage, feature extraction stage, classification stage, and testing stage. In pre-processing stage, the digital image processing methods, which are noise reduction, contrast enhancement, thresholding, and morphological and logical processes. In feature extraction stage, the invariant moments of pre-processed parasite images are calculated. Finally, in classification stage, the multi-class support vector machine (MCSVM) classifier is used for classification of features extracted feature extraction stage. We used MATLAB software for estimating the success classification rate of proposed approach in this study. For this aim, proposed approach was tested by using test data. At end of test, 97.70% overall success rates were obtained.

A novel approach for digital radio signal classification: Wavelet packet energy-multiclass support vector machine (WPE-MSVM)

In this study, a novel application of wavelet packet energy-multicass support vector machine (WPE-MSVM) is proposed to perform automatic modulation classification of digital radio signals. In this approach, first, the discrete wavelet packet transforms (DWPTs) of digital modulated radio signal types are performed. Second, the wavelet packet energies of these DWPTs are calculated. Third, these wavelet packet energy features are given to inputs of multiclass support vector machine (MSVM) classifier. Fourth, test data is given to inputs of MSVM classifier for evaluating the classification performance of this proposed classification approach. Here, db2, db3, db4, db5, db8, sym2, sym3, sym5, sym7, sym8, bior1.3, bior2.2, bior2.8, coif1 and coif5 wavelet packet decomposition filters are separately used for DWPT of these digital modulated radio signals, respectively. Thus, performance comparisons of these wavelet packet decomposition filters for digital radio signal classification are performed by using wavelet packet energy features. The digital radio signal types used in this study are 9 types, which are ASK-2, ASK-4, ASK-8, FSK-2, FSK-4, FSK-8, PSK-2, PSK-4 and PSK-8. These experimental studies are realized by using total 2250 digital modulated signals for these digital radio signal types. The rate of mean correct classification is about 90% for the sample digital modulated signals.

An expert system based on fuzzy entropy for automatic threshold selection in image processing

In pattern recognition and image processing, the selection of appropriate threshold is a very significant issue. Especially, the selecting gray-level thresholds is a critical issue for many pattern recognition applications. Here, the maximum fuzzy entropy and fuzzy c-partition methods are used for the aim of the gray-level automatic threshold selection method. The fuzzy theory has been successfully applied to many areas, such as image processing, pattern recognition, computer vision, medicine, control, etc. The images have some fuzziness in nature. In this study, expert maximum fuzzy-Sure entropy (EMFSE) method for the maximum fuzzy entropy and fuzzy c-partition processes in automatic threshold selection is proposed. The experimental studies were conducted on many images by testing maximum fuzzy-Sure entropy against maximum fuzzy-Shannon entropy (MFSHE), maximum fuzzy-Havrada and Charvat entropy (MFHCE) methods for selecting optimum 2-level threshold value, respectively. The obtained experimental results show that the used MFSE method is superior to other MFSHE and MFHCE methods on selecting the 2-level threshold value automatically and effectively.

An automatic diagnosis system based on thyroid gland: ADSTG

In this paper, the automatic diagnosis system based on thyroid gland (ADSTG) method is introduced for diagnosis of thyroid disease. The structure of this ADSTG diagnosis system for thyroid diseases contains three stages. In first stage, the feature reduction is performed by using Principle Component Analysis (PCA) method. In second stage, the classification by using Least Square Support Vector Machine (LS-SVM) classifier. In third stage, the performance evaluation of this ADSTG method for diagnosis of thyroid disease is estimated by using classification accuracy, k-fold cross-validation, and confusion matrix methods respectively. The classification accuracy of this ADSTG diagnosis system for thyroid diseases was obtained about 97.67%.

The performance comparison of discrete wavelet neural network and discrete wavelet adaptive network based fuzzy inference system for digital modulation recognition

In this paper, a new discrete wavelet neural network (DWNN) and discrete wavelet adaptive network based fuzzy inference system (DWANFIS) methods are offered for automatic digital modulation recognition (ADMR) and the performance comparison between these new DWNN and DWANFIS intelligent systems are performed by using bior1.3, bior2.2, bior2.8, bior3.5, bior6.8, coif1, coif2, coif3, coif4, coif5, db3, db5, db8, db10, sym2, sym3, sym5, sym7, and sym8 wavelet decomposition filters, respectively. Moreover in this study, discrete wavelet transform (DWT) and adaptive wavelet entropy are used in feature extraction stages of these intelligent systems. The digital modulation types used in this study are ASK2, ASK4, ASK8, FSK2, FSK4, FSK8, PSK2, PSK4, and PSK8. Here, mean correct recognition rates for digital modulation recognition were obtained 96.51% and 90.24% by using DWNN and DWANFIS intelligent systems, respectively.