Browsing by Subject "Principal component analysis"

Now showing 1 - 11 of 11

Embargo
A principal component model to identify Turkish soundscapes’ affective attributes based on a corpus-driven approach
(Elsevier, 2023-06-30) Yılmazer, Semiha; Fasllija, Ela; Alimadhi, Enkela; Şahin, Zekiye; Mercan, Elif; Dalirnaghadeh, Donya
This study focused on achieving linguistic and culturally appropriate equivalents of Turkish soundscape attributes present in ISO 12913–3 by incorporating a Corpus-Driven Approach (CDA). A two-phase experiment was set up to find Turkish equivalents of affective quality attributes. The first phase consisted of the formation of a Corpus. An online questionnaire was prepared and sent to 196 native Turkish speakers from all around Türkiye to define adjectives. The second phase of the experiment was performed in a listening room. For this purpose, twenty-four binaural sound recordings were collected from seven public spaces. Afterward, forty individuals evaluated the recordings by using the attributes from Phase 1. The perceptual dimensions were obtained from the generated corpus in Turkish based on a rating scale by applying the Principal Component Analysis (PCA). Results indicated a two-dimensional model with two main components, Pleasantness and Eventfulness. Each component is associated with a main orthogonal axis denoted by ‘annoying-comfortable’ and ‘dynamic-uneventful,’ respectively. This circular organization of soundscape attributes is supported by two derived axes, namely ‘chaotic-calm’ and ‘monotonous-enjoyable’, rotated 45°on the same plane. Additionally, by using Spearman's rank correlation coefficient, sixty-four different bipolar adjective pairs were found. The adjective pairs showed that the highest correlations are mainly on the pleasant-unpleasant continuum, namely Component 1 of PCA. The collected data were also analyzed using Agglomerative Hierarchical Cluster analysis with the Ward method in R programming language to cluster the adjectives. The results inferred that there are four top-level categories. From the first to the fourth level, categories consisted of pleasant, uneventful, eventful, and annoying adjectives, respectively. Moreover, the terms grouped on the first cluster found their dichotomous on the fourth cluster, while maintaining the same relationship in the pleasant-unpleasant continuum.
Open Access
Colour and design: from natural patterns to monochrome compositions
(Elsevier, 2011-03) Olguntürk, N.; Demirkan, H.
There is no doubt that nature provides endless inspiration to the world of design. In order to explore the role of colour in design, forty-two people were asked to first choose a pattern from nature, then to abstract this pattern into geometric shapes and finally to colour this pattern. All work done by the participants were statistically analysed to find out the effect of colour on design. Findings of the study suggest that colour in a pattern is the first principal component of design as a unifier whereas this is replaced with the number of shapes in one group in black and white patterns.
Open Access
Eye tracking using markov models
(IEEE, 2004) Bağcı, A. M.; Ansari, R.; Khokhar, A.; Çetin, A. Enis
We propose an eye detection and tracking method based on color and geometrical features of the human face using a monocular camera. In this method a decision is made on whether the eyes are closed or not and, using a Markov chain framework to model temporal evolution, the subject's gaze is determined. The method can successfully track facial features even while the head assumes various poses, so long as the nostrils are visible to the camera. We compare our method with recently proposed techniques and results show that it provides more accurate tracking and robustness to variations in view of the face. A procedure for detecting tracking errors is employed to recover the loss of feature points in case of occlusion or very fast head movement. The method may be used in monitoring a driver's alertness and detecting drowsiness, and also in applications requiring non-contact human computer interaction.
Open Access
Fast insect damage detection in wheat kernels using transmittance images
(IEEE, 2004-07) Çataltepe, Z.; Pearson, T.; Cetin, A. Enis
We used transmittance images and different learning algorithms to classify insect damaged and un-damaged wheat kernels. Using the histogram of the pixels of the wheat images as the feature, and the linear model as the learning algorithm, we achieved a False Positive Rate (1-specificity) of 0.12 at the True Positive Rate (sensitivity) of 0.8 and an Area Under the ROC Curve (AUC) of 0.90 ± 0.02. Combining the linear model and a Radial Basis Function Network in a committee resulted in a FP Rate of 0.09 at the TP Rate of 0.8 and an AUC of 0.93 ± 0.03.
Open Access
Human activity recognition using inertial/magnetic sensor units
(Springer, Berlin, Heidelberg, 2010) Altun, Kerem; Barshan, Billur
This paper provides a comparative study on the different techniques of classifying human activities that are performed using body-worn miniature inertial and magnetic sensors. The classification techniques implemented and compared in this study are: Bayesian decision making (BDM), the least-squares method (LSM), the k-nearest neighbor algorithm (k-NN), dynamic time warping (DTW), support vector machines (SVM), and artificial neural networks (ANN). Daily and sports activities are classified using five sensor units worn by eight subjects on the chest, the arms, and the legs. Each sensor unit comprises a triaxial gyroscope, a triaxial accelerometer, and a triaxial magnetometer. Principal component analysis (PCA) and sequential forward feature selection (SFFS) methods are employed for feature reduction. For a small number of features, SFFS demonstrates better performance and should be preferable especially in real-time applications. The classifiers are validated using different cross-validation techniques. Among the different classifiers we have considered, BDM results in the highest correct classification rate with relatively small computational cost. © 2010 Springer-Verlag Berlin Heidelberg.
Open Access
Kentsel yapıların biçimbilimsel bölütlenmesi
(IEEE, 2007-06) Akçay, H. Gökhan; Aksoy, Selim
Yüksek çözünürlükteki uzaktan algılamalı uydu görüntülerinde bölütleme kent uygulamalarında önemli bir problemdir çünkü elde edilen bölütlemelerle sınıflandırma için piksel tabanlı spektral bilginin yanında uzamsal ve yapısal bilgiler elde edilebilir. Bu bildiride, biçimbilimsel işlemlerle çıkarılan yapısal bilgi ve ana bileşenler analizi ile özetlenen spektral bilgi kullanılarak gürültüden etkilenmeyen bölütler elde eden bir yöntem sunduk. Yapılan deneyler yöntemin görüntü üzerinde komşuluk bilgisini ve spektral bilgiyi beraber kullanmayan başka bir yönteme göre daha düzgün ve anlamlı yapılar bulduğunu göstermiştir. Automatic segmentation of high-resolution remote sensing imagery is an important problem in urban applications because the resulting segmentations can provide valuable spatial and structural information that are complementary to pixel-based spectral information in classification. We present a method that combines structural information extracted by morphological processing with spectral information summarized using principal components analysis to produce precise segmentations that are also robust to noise. The experiments show that the method is able to detect structures in the image which are more precise and more meaningful than the structures detected by another approach that does not make strong use of neighborhood and spectral information.
Open Access
Man-made object classification in SAR images using 2-D cepstrum
(IEEE, 2009-05) Eryildirim, A.; Çetin, A. Enis
In this paper, a novel descriptive feature parameter extraction method from Synthetic Aperture Radar (SAR) images is proposed. The new method is based on the two-dimensional (2-D) real cepstrum. This novel 2-D cepstrum method is compared with principal component analysis (PCA) method by testing over the MSTAR image database. The extracted features are classified using Support Vector Machine (SVM). We demonstrate that discrimination of natural background (clutter) and man-made objects (metal objects) in SAR imagery is possible using the 2-D cepstrum feature parameters. In addition, the computational cost of the cepstrum method is lower than the PCA method. Experimental results are presented. ©2009 IEEE.
Open Access
Recognizing daily and sports activities in two open source machine learning environments using body-worn sensor units
(Oxford University Press, 2014-11) Barshan, B.; Yüksek, M. C.
This study provides a comparative assessment on the different techniques of classifying human activities performed while wearing inertial and magnetic sensor units on the chest, arms and legs. The gyroscope, accelerometer and the magnetometer in each unit are tri-axial. Naive Bayesian classifier, artificial neural networks (ANNs), dissimilarity-based classifier, three types of decision trees, Gaussian mixture models (GMMs) and support vector machines (SVMs) are considered. A feature set extracted from the raw sensor data using principal component analysis is used for classification. Three different cross-validation techniques are employed to validate the classifiers. A performance comparison of the classifiers is provided in terms of their correct differentiation rates, confusion matrices and computational cost. The highest correct differentiation rates are achieved with ANNs (99.2%), SVMs (99.2%) and a GMM (99.1%). GMMs may be preferable because of their lower computational requirements. Regarding the position of sensor units on the body, those worn on the legs are the most informative. Comparing the different sensor modalities indicates that if only a single sensor type is used, the highest classification rates are achieved with magnetometers, followed by accelerometers and gyroscopes. The study also provides a comparison between two commonly used open source machine learning environments (WEKA and PRTools) in terms of their functionality, manageability, classifier performance and execution times. © 2013 © The British Computer Society 2013. All rights reserved.
Open Access
Strain-and region-specific gene expression profiles in mouse brain in response to chronic nicotine treatment
(Wiley-Blackwell Publishing, 2008) Wang, J.; Gutala, R.; Hwang, Y. Y.; Kim J. -M.; Konu, O.; Ma, J. Z.; Li, M. D.
A pathway-focused complementary DNA microarray and gene ontology analysis were used to investigate gene expression profiles in the amygdala, hippocampus, nucleus accumbens, prefrontal cortex (PFC) and ventral tegmental area of C3H/HeJ and C57BL/6J mice receiving nicotine in drinking water (100 μg/ml in 2% saccharin for 2 weeks). A balanced experimental design and rigorous statistical analysis have led to the identification of 3.5-22.1% and 4.1-14.3% of the 638 sequence-verified genes as significantly modulated in the aforementioned brain regions of the C3H/HeJ and C57BL/6J strains, respectively. Comparisons of differential expression among brain tissues showed that only a small number of genes were altered in multiple brain regions, suggesting presence of a brain region-specific transcriptional response to nicotine. Subsequent principal component analysis and Expression Analysis Systematic Explorer analysis showed significant enrichment of biological processes both in C3H/HeJ and C57BL/6J mice, i.e. cell cycle/proliferation, organogenesis and transmission of nerve impulse. Finally, we verified the observed changes in expression using real-time reverse transcriptase polymerase chain reaction for six representative genes in the PFC region, providing an independent replication of our microarray results. Together, this report represents the first comprehensive gene expression profiling investigation of the changes caused by nicotine in brain tissues of the two mouse strains known to exhibit differential behavioral and physiological responses to nicotine.
Open Access
Target detection and classification in SAR images using region covariance and co-difference
(SPIE, 2009-04) Duman, Kaan; Eryıldırım, Abdulkadir; Çetin, A. Enis
In this paper, a novel descriptive feature parameter extraction method from synthetic aperture radar (SAR) images is proposed. The new approach is based on region covariance (RC) method which involves the computation of a covariance matrix whose entries are used in target detection and classification. In addition the region co-difference matrix is also introduced. Experimental results of object detection in MSTAR (moving and stationary target recognition) database are presented. The RC and region co-difference method delivers high detection accuracy and low false alarm rates. It is also experimentally observed that these methods produce better results than the commonly used principal component analysis (PCA) method when they are used with different distance metrics introduced. © 2009 SPIE.
Open Access
Wheat and hazelnut inspection with impact acoustics time-frequency patterns
(ASABE, 2007-06) İnce, N. F.; Onaran, İbrahim; Tewfik, A. H.; Kalkan, H.; Pearson, T.; Çetin, A. Enis; Yardimci, Y.
Kernel damage caused by insects and fungi is one of the most common reason for poor flour quality. Cracked hazelnut shells are prone to infection by cancer producing mold. We propose a new adaptive time-frequency classification procedure for detecting cracked hazelnut shells and damaged wheat kernels using impact acoustic emissions recorded by dropping wheat kernels or hazelnut shells on a steel plate. The proposed algorithm is based on a flexible local discriminant bases (F-LDB) procedure. The F-LDB method combines local cosine packet analysis and a frequency axis clustering approach which supports individual time and frequency band adaptation. Discriminant features are extracted from the adaptively segmented acoustic signal, sorted according to a Fisher class separability criterion, post processed by principal component analysis and fed to linear discriminant. We describe experimental results that establish the superior performance of the proposed approach when compared with prior techniques reported in the literature or used in the field. Our approach achieved classification accuracy in paired separation of undamaged wheat kernels from IDK, Pupae and Scab damaged kernels with 96%, 82% and 94%. For hazelnuts the accuracy was 97.1%.