Bülent Bolat

Active learning for probabilistic neural networks

In many neural network applications, the selection of best training set to represent the entire sample space is one of the most important problems. Active learning algorithms in the literature for neural networks are not appropriate for Probabilistic Neural Networks (PNN). In this paper, a new active learning method is proposed for PNN. The method was applied to several benchmark problems.

Classification of Parkinson's disease by using voice measurements

In this study, a new approach has been presented to classify Parkinsons disease (PD). In order to discriminate healthy people from the PD patients, several measurements extracted from sound samples of 31 people, 23 with PD, have been applied to four different classifiers. In order to classify the subject as PD patient or healthy, a probabilistic neural network (PNN), a generalised regression neural network (GRNN), a support vector machine and a k-nearest neighbour have been carried out. Half of the dataset are used for training, remaining data are used for testing in order to determine the performance of the classifiers. In each classification process two-fold cross validation method is utilised to determine which subset represents the entire dataset. It is shown that reasonable results can be obtained by following the proposed methods.

A CAD of fully automated colonic polyp detection for contrasted and non-contrasted CT scans.

PurposeComputer-aided detection (CAD) systems are developed to help radiologists detect colonic polyps over CT scans. It is possible to reduce the detection time and increase the detection accuracy rates by using CAD systems. In this paper, we aimed to develop a fully integrated CAD system for automated detection of polyps that yields a high polyp detection rate with a reasonable number of false positives.MethodsThe proposed CAD system is a multistage implementation whose main components are: automatic colon segmentation, candidate detection, feature extraction and classification. The first element of the algorithm includes a discrete segmentation for both air and fluid regions. Colon-air regions were determined based on adaptive thresholding, and the volume/length measure was used to detect air regions. To extract the colon-fluid regions, a rule-based connectivity test was used to detect the regions belong to the colon. Potential polyp candidates were detected based on the 3D Laplacian of Gaussian filter. The geometrical features were used to reduce false-positive detections. A 2D projection image was generated to extract discriminative features as the inputs of an artificial neural network classifier.ResultsOur CAD system performs at 100% sensitivity for polyps larger than 9 mm, 95.83% sensitivity for polyps 6–10 mm and 85.71% sensitivity for polyps smaller than 6 mm with 5.3 false positives per dataset. Also, clinically relevant polyps (

Speeh/music classification by using statistical neural networks

\ge

Feature extraction and classification of neuromuscular diseases using scanning EMG

≥6 mm) were identified with 96.67% sensitivity at 1.12 FP/dataset.ConclusionsTo the best of our knowledge, the novel polyp candidate detection system which determines polyp candidates with LoG filters is one of the main contributions. We also propose a new 2D projection image calculation scheme to determine the distinctive features. We believe that our CAD system is highly effective for assisting radiologist interpreting CT.

Detection of pedestrians from still images

In this work, Classical Turkish Music songs are classified into six makams. Makam is a modal framework for melodic development in Classical Turkish Music. The effect of the sound clip length on the system performance was also evaluated. The Mel Frequency Cepstral Coefficients (MFCC) were used as features. Obtained data were classified by using Probabilistic Neural Network. The best correct recognition ratio was obtained as 89,4% by using a clip length of 6 s.

Pedestrian detection from still images

This paper represents a framework for speech/music classification by using statistical neural networks. Zero crossing rate, root mean square power and spectral centroid were used as features. A dataset including 150 audio instances was labeled manually and 105 of them were used to train different networks, which are the probabilistic neural network (PNN), the generalised regression neural network (GRNN) and the radial basis functions (RBF). The remainder of the dataset was used as test item. Training and test performance of these three network types were discussed.

Determining MUAP activity corridor in scanning EMG recordings

Auscultation and analysing of lung sound is widely used in clinical area for diagnosis of lung diseases. Due to the non-stationary nature of lung sounds conventional frequency analysis technique is not a successful method for respiratory sound analysis. In this paper, classification of normal and abnormal lung sound using wavelet coefficient intended. Respiratory sounds are decomposed into the frequency subbands using wavelet transform and a set of statistical features are inspected from the sub-bands. Then, lung sounds classified as normal and abnormal using these statistical features. Artificial neural network and support vector machine are used for classification process.