Browsing by Author "Özkan, H."

Now showing 1 - 9 of 9

Open Access
Adaptive hierarchical space partitioning for online classification
(IEEE, 2016) Kılıç, O. Fatih; Vanlı, N. D.; Özkan, H.; Delibalta, İ.; Kozat, Süleyman Serdar
We propose an online algorithm for supervised learning with strong performance guarantees under the empirical zero-one loss. The proposed method adaptively partitions the feature space in a hierarchical manner and generates a powerful finite combination of basic models. This provides algorithm to obtain a strong classification method which enables it to create a linear piecewise classifier model that can work well under highly non-linear complex data. The introduced algorithm also have scalable computational complexity that scales linearly with dimension of the feature space, depth of the partitioning and number of processed data. Through experiments we show that the introduced algorithm outperforms the state-of-the-art ensemble techniques over various well-known machine learning data sets.
Open Access
Boosted adaptive filters
(Elsevier, 2018) Kari, Dariush; Mirza, Ali H.; Khan, Farhan; Özkan, H.; Kozat, Süleyman Serdar
We introduce the boosting notion of machine learning to the adaptive signal processing literature. In our framework, we have several adaptive filtering algorithms, i.e., the weak learners, that run in parallel on a common task such as equalization, classification, regression or filtering. We specifically provide theoretical bounds for the performance improvement of our proposed algorithms over the conventional adaptive filtering methods under some widely used statistical assumptions. We demonstrate an intrinsic relationship, in terms of boosting, between the adaptive mixture-of-experts and data reuse algorithms. Additionally, we introduce a boosting algorithm based on random updates that is significantly faster than the conventional boosting methods and other variants of our proposed algorithms while achieving an enhanced performance gain. Hence, the random updates method is specifically applicable to the fast and high dimensional streaming data. Specifically, we investigate Recursive Least Square-based and Least Mean Square-based linear and piecewise-linear regression algorithms in a mixture-of-experts setting and provide several variants of these well-known adaptation methods. Furthermore, we provide theoretical bounds for the computational complexity of our proposed algorithms. We demonstrate substantial performance gains in terms of mean squared error over the base learners through an extensive set of benchmark real data sets and simulated examples.
Open Access
Efficient NP tests for anomaly detection over birth-death type DTMCs
(Springer New York LLC, 2018) Özkan, H.; Özkan, F.; Delibalta, I.; Kozat, Süleyman S.
We propose computationally highly efficient Neyman-Pearson (NP) tests for anomaly detection over birth-death type discrete time Markov chains. Instead of relying on extensive Monte Carlo simulations (as in the case of the baseline NP), we directly approximate the log-likelihood density to match the desired false alarm rate; and therefore obtain our efficient implementations. The proposed algorithms are appropriate for processing large scale data in online applications with real time false alarm rate controllability. Since we do not require parameter tuning, our algorithms are also adaptive to non-stationarity in the data source. In our experiments, the proposed tests demonstrate superior detection power compared to the baseline NP while nearly achieving the desired rates with negligible computational resources.
Open Access
Nonlinear regression via incremental decision trees
(Elsevier, 2019) Vanlı, N.; Sayın, M.; Neyshabouri, Mohammadreza Mohaghegh; Özkan, H.; Kozat, Süleyman S.
We study sequential nonlinear regression and introduce an online algorithm that elegantly mitigates, via an adaptively incremental hierarchical structure, convergence and undertraining issues of conventional nonlinear regression methods. Particularly, we present a piecewise linear (or nonlinear) regression algorithm that partitions the regressor space and learns a linear model at each region to combine. Unlike the conventional approaches, our algorithm effectively learns the optimal regressor space partition with the desired complexity in a completely sequential and data driven manner. Our algorithm sequentially and asymptotically achieves the performance of the optimal twice differentiable regression function for any data sequence without any statistical assumptions. The introduced algorithm can be efficiently implemented with a computational complexity that is only logarithmic in the length of data. In our experiments, we demonstrate significant gains for the well-known benchmark real data sets when compared to the state-of-the-art techniques.
Open Access
Parametrik olmayan yoğunluk tahmincileri ile ardışık anomali tespiti
(IEEE, 2019-04) Kerpiççi, Mine; Kozat, Süleyman S.; Özkan, H.
Bu bildiride, gözlemlenen verideki anomalileri, gözetimsiz bir çerçevede, iki aşamalı yöntemle bulmak için anomali tespit algoritması tanıtılmıştır. İlk aşamada, ardışık olarak gözlemlenen verinin yoğunluğu çekirdek temelli özgün bir yöntemle tahmin edilmektedir. Bu amaçla, gözlem alanı bölünmekte ve her bölgede parametrik olmayan Çekirdek Yoğunluk Tahmincisi (ÇYT) veri dağılımına dair hiçbir varsayımda bulunulmadan kullanılmaktadır. Sonra, yoğunluk tahmini eşik değeriyle karşılaştırılarak verinin anomali olup olmadığına karar verilmektedir. Ayrıca, çekirdek temelli yöntemlerdeki bant genişliği seçimi problemi de verimli bir şekilde çözülmektedir. Bu amaçla, her bir bölgeye çekirdek bant genişliği seti atanmakta ve her tahmincinin ait olduğu bölgeye göre en iyi bant genişliği seçeneğine zamanla ulaşması sağlanmaktadır. Sayısal örneklerde, tanıtılan algoritmanın literatürde sıklıkla kullanılan anomali tespit metodlarına göre yüksek performans artışı elde ettiği gösterilmektedir.
Open Access
Resonant Raman scattering near the free-to-bound transition in undoped p-GaSe
(Wiley, 2001) Gasanly, N. M.; Aydınlı, A.; Özkan, H.
Raman spectra of GaSe layered crystal have been measured using a He-Ne laser and temperature tuning the free-to-bound gap in the range 10-290 K. Resonance enhancement of E’’(2) mode has been observed for both incident and scattered photon energies equal to the free-to-bound transition energy.
Open Access
Temperature dependence of the first-order Raman scattering in GaS layered crystals
(Pergamon Press, 2000) Gasanly, N. M.; Aydınlı, A.; Özkan, H.; Kocabaş, C.
The temperature dependence (15-293 K) of the six Raman-active mode frequencies and linewidths in gallium sulfide has been measured in the frequency range from 15 to 380 cm-1. We observed softening and broadening of the optical phonon lines with increasing temperature. Comparison between the experimental data and theories of the shift and broadening of the interlayer and intralayer phonon lines during the heating of the crystal showed that the experimental dependencies can be explained by the contributions from thermal expansion and lattice anharmonicity. The pure-temperature contribution (phonon-phonon coupling) is due to three- and four-phonon processes.
Open Access
Temperature dependence of the Raman-active phonon frequencies in indium sulfide
(Pergamon Press, 1999) Gasanly, N. M.; Özkan, H.; Aydınlı, Atilla; Yilmaz, I.
The temperature dependence of the Raman-active mode frequencies in indium sulfide was measured in the range from 10 to 300 K. The analysis of the temperature dependence of the A g intralayer optical modes show that Raman frequency shift results from the change of harmonic frequency with volume expansion and anharmonic coupling to phonons of other branches. The pure-temperature contribution (phonon-phonon coupling) is due to three- and four-phonon processes.
Open Access
Temperature-dependent Raman scattering spectra of ε-GaSe layered crystal
(Elsevier Science, 2002) Gasanly, N. M.; Aydnl, A.; Özkan, H.; Kocabaş, C.
The temperature dependencies (15-300 K) of seven Raman-active mode frequencies and linewidths in layered gallium selenide have been measured in the frequency range from 10 to 320 cm-1. We observed softening and broadening of the optical phonon lines with increasing temperature. Comparison between the experimental data and theories of the shift and broadening of the intralayer phonon lines during heating of the crystal showed that the experimental dependencies can be explained by the contributions from thermal expansion, lattice anharmonicity and crystal disorder. The pure-temperature contribution (phonon-phonon coupling) is due to three-phonon processes. Moreover, it was established that the effect of crystal disorder on the linewidth broadening of TO mode is stronger than that of LO mode.