Arif Selcuk Ogrenci
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Arif Selcuk Ogrenci has a total of 50 co-authors. Here are the top co-authors by number of shared publications:
- Ayse Humeyra Bilge
Kadir Has University
5 shared publications - Önder Pekcan
Faculty of Engineering and Natural Sciences, Kadir Has University, Istanbul, Turkey
4 shared publications Erdal Panayirci
3 shared publications- Gunhan Dundar
Department of Electrical and Electronics Engineering, Bogazici University, Bebek, Istanbul, Turkey
3 shared publications - Mehmet Nafiz Aydin
Kadir Has University
2 shared publications - Mhd Tahssin Altabbaa
Department of Electrical and Electronics Engineering, Istanbul Yeni Yüzyıl University, Istanbul, Turkey
2 shared publications
Latest Publications
- 1 February 2023
- Published by Taylor & Francis in Journal of Macromolecular Science, Part B
- Vol. 62 (2) , 66-84
Epidemic Models for Chemical Gel Phase Transformation: Effect of Optical Measurements with Different Wavelengths
In our previous work we characterized the sol-gel transition of chemical and physical gels in terms of epidemic models which describe the spread of a disease in a society. We have shown that gelation in chemical gels can be represented by the Susceptible-Infected-Recovered (SIR) model, while the gelation mechanism of physical gels can be represented by the Susceptible-Infected-Susceptible (SIS) model. These studies were based on optical measurements at a single wavelength. In the present work we studied the gelation mechanism for chemical gels obtained from experiments that have been carried out with acrylamide (AAm) and N,N0-methylenebis (acrylamide) (Bis) as base materials for chemical gel formation in water and with cross-linker. The transmitted light intensity was measured for various wavelengths from 440nm to 690nm during the gelation process. The data are modeled by the epidemic SIR process. The results indicate that the gelation processes obtained from the same Bis and water contents observed at different wavelengths obey different SIR models. This observation is explained by the scattering of light of different wavelengths depending on the sizes of the micro-gels formed.
- 1 February 2023
- Published by Taylor & Francis in Journal of Macromolecular Science, Part B
- Vol. 62 (2) , 66-84
Epidemic Models for Chemical Gel Phase Transformation: Effect of Optical Measurements with Different Wavelengths
In our previous work we characterized the sol-gel transition of chemical and physical gels in terms of epidemic models which describe the spread of a disease in a society. We have shown that gelation in chemical gels can be represented by the Susceptible-Infected-Recovered (SIR) model, while the gelation mechanism of physical gels can be represented by the Susceptible-Infected-Susceptible (SIS) model. These studies were based on optical measurements at a single wavelength. In the present work we studied the gelation mechanism for chemical gels obtained from experiments that have been carried out with acrylamide (AAm) and N,N0-methylenebis (acrylamide) (Bis) as base materials for chemical gel formation in water and with cross-linker. The transmitted light intensity was measured for various wavelengths from 440nm to 690nm during the gelation process. The data are modeled by the epidemic SIR process. The results indicate that the gelation processes obtained from the same Bis and water contents observed at different wavelengths obey different SIR models. This observation is explained by the scattering of light of different wavelengths depending on the sizes of the micro-gels formed.
- 23 September 2021
- Published by Springer Nature in Environmental Science and Pollution Research
- Vol. 29 (8) , 11920-11938
Air quality prediction using CNN+LSTM-based hybrid deep learning architecture
Air pollution prediction based on variables in environmental monitoring data gains further importance with increasing concerns about climate change and the sustainability of cities. Modeling of the complex relationships between these variables by sophisticated methods in machine learning is a promising field. The objectives of this work are to develop a supervised model for the prediction of air pollution by using real sensor data and to transfer the model between cities. The combination of a convolutional neural network and a long short-term memory deep neural network model was proposed to predict the concentration of air pollutants in multiple locations of a city by using spatial-temporal relationships. Two approaches have been adopted: the univariate model contains the information of one pollutant whereas the multivariate model contains the information of all pollutants and meteorology data for prediction. The study was carried out for different pollutants which are in the publicly available data of the cities of Barcelona, Kocaeli, and İstanbul. The hyperparameters of the model (filter, frame, and batch sizes; number of convolutional/LSTM layers and hidden units; learning rate; and parameters for sample selection, pooling, and validation) were tuned to determine the architecture that achieved the lowest test error. The proposed model improved the prediction performance (measured by the root mean square error) by 11–53% for particulate matter, 20–31% for ozone, 9–47% for nitrogenoxides, and 18–46% for sulfurdioxide with respect to the 1-hidden layer long short-term memory networks utilized in the literature. The multivariate model without using meteorological data revealed the best results. Regarding transfer learning, the network weights were transferred from the source city to the target city. The model has more accurate prediction performance with the transfer of the network from Kocaeli to İstanbul as those neighbor cities have similar air pollution and meteorological characteristics.
- 23 September 2021
- Published by Springer Nature in Environmental Science and Pollution Research
- Vol. 29 (8) , 11920-11938
Air quality prediction using CNN+LSTM-based hybrid deep learning architecture
Air pollution prediction based on variables in environmental monitoring data gains further importance with increasing concerns about climate change and the sustainability of cities. Modeling of the complex relationships between these variables by sophisticated methods in machine learning is a promising field. The objectives of this work are to develop a supervised model for the prediction of air pollution by using real sensor data and to transfer the model between cities. The combination of a convolutional neural network and a long short-term memory deep neural network model was proposed to predict the concentration of air pollutants in multiple locations of a city by using spatial-temporal relationships. Two approaches have been adopted: the univariate model contains the information of one pollutant whereas the multivariate model contains the information of all pollutants and meteorology data for prediction. The study was carried out for different pollutants which are in the publicly available data of the cities of Barcelona, Kocaeli, and İstanbul. The hyperparameters of the model (filter, frame, and batch sizes; number of convolutional/LSTM layers and hidden units; learning rate; and parameters for sample selection, pooling, and validation) were tuned to determine the architecture that achieved the lowest test error. The proposed model improved the prediction performance (measured by the root mean square error) by 11–53% for particulate matter, 20–31% for ozone, 9–47% for nitrogenoxides, and 18–46% for sulfurdioxide with respect to the 1-hidden layer long short-term memory networks utilized in the literature. The multivariate model without using meteorological data revealed the best results. Regarding transfer learning, the network weights were transferred from the source city to the target city. The model has more accurate prediction performance with the transfer of the network from Kocaeli to İstanbul as those neighbor cities have similar air pollution and meteorological characteristics.
- Open Access
- 7 September 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 8, 164155-164177
Unsupervised Anomaly Detection in Multivariate Spatio-Temporal Data Using Deep Learning: Early Detection of COVID-19 Outbreak in Italy
Unsupervised anomaly detection for spatio-temporal data has extensive use in a wide variety of applications such as earth science, traffic monitoring, fraud and disease outbreak detection. Most real-world time series data have a spatial dimension as an additional context which is often expressed in terms of coordinates of the region of interest (such as latitude - longitude information). However, existing techniques are limited to handle spatial and temporal contextual attributes in an integrated and meaningful way considering both spatial and temporal dependency between observations. In this paper, a hybrid deep learning framework is proposed to solve the unsupervised anomaly detection problem in multivariate spatio-temporal data. The proposed framework works with unlabeled data and no prior knowledge about anomalies are assumed. As a case study, we use the public COVID-19 data provided by the Italian Department of Civil Protection. Northern Italy regions’ COVID-19 data are used to train the framework; and then any abnormal trends or upswings in COVID-19 data of central and southern Italian regions are detected. The proposed framework detects early signals of the COVID-19 outbreak in test regions based on the reconstruction error. For performance comparison, we perform a detailed evaluation of 15 algorithms on the COVID-19 Italy dataset including the state-of-the-art deep learning architectures. Experimental results show that our framework shows significant improvement on unsupervised anomaly detection performance even in data scarce and high contamination ratio scenarios (where the ratio of anomalies in the data set is more than 5%). It achieves the earliest detection of COVID-19 outbreak and shows better performance on tracking the peaks of the COVID-19 pandemic in test regions. As the timeliness of detection is quite important in the fight against any outbreak, our framework provides useful insight to suppress the resurgence of local novel coronavirus outbreaks as early as possible.
- 7 September 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 8, 164155-164177
Open AccessUnsupervised Anomaly Detection in Multivariate Spatio-Temporal Data Using Deep Learning: Early Detection of COVID-19 Outbreak in Italy
Unsupervised anomaly detection for spatio-temporal data has extensive use in a wide variety of applications such as earth science, traffic monitoring, fraud and disease outbreak detection. Most real-world time series data have a spatial dimension as an additional context which is often expressed in terms of coordinates of the region of interest (such as latitude - longitude information). However, existing techniques are limited to handle spatial and temporal contextual attributes in an integrated and meaningful way considering both spatial and temporal dependency between observations. In this paper, a hybrid deep learning framework is proposed to solve the unsupervised anomaly detection problem in multivariate spatio-temporal data. The proposed framework works with unlabeled data and no prior knowledge about anomalies are assumed. As a case study, we use the public COVID-19 data provided by the Italian Department of Civil Protection. Northern Italy regions’ COVID-19 data are used to train the framework; and then any abnormal trends or upswings in COVID-19 data of central and southern Italian regions are detected. The proposed framework detects early signals of the COVID-19 outbreak in test regions based on the reconstruction error. For performance comparison, we perform a detailed evaluation of 15 algorithms on the COVID-19 Italy dataset including the state-of-the-art deep learning architectures. Experimental results show that our framework shows significant improvement on unsupervised anomaly detection performance even in data scarce and high contamination ratio scenarios (where the ratio of anomalies in the data set is more than 5%). It achieves the earliest detection of COVID-19 outbreak and shows better performance on tracking the peaks of the COVID-19 pandemic in test regions. As the timeliness of detection is quite important in the fight against any outbreak, our framework provides useful insight to suppress the resurgence of local novel coronavirus outbreaks as early as possible.
- Open Access
- 28 July 2020
- Published by MDPI AG in Applied Sciences
- Vol. 10 (15) , 5191
A Hybrid Deep Learning Framework for Unsupervised Anomaly Detection in Multivariate Spatio-Temporal Data
Multivariate time-series data with a contextual spatial attribute have extensive use for finding anomalous patterns in a wide variety of application domains such as earth science, hurricane tracking, fraud, and disease outbreak detection. In most settings, spatial context is often expressed in terms of ZIP code or region coordinates such as latitude and longitude. However, traditional anomaly detection techniques cannot handle more than one contextual attribute in a unified way. In this paper, a new hybrid approach based on deep learning is proposed to solve the anomaly detection problem in multivariate spatio-temporal dataset. It works under the assumption that no prior knowledge about the dataset and anomalies are available. The architecture of the proposed hybrid framework is based on an autoencoder scheme, and it is more efficient in extracting features from the spatio-temporal multivariate datasets compared to the traditional spatio-temporal anomaly detection techniques. We conducted extensive experiments using buoy data of 2005 from National Data Buoy Center and Hurricane Katrina as ground truth. Experiments demonstrate that the proposed model achieves more than 10% improvement in accuracy over the methods used in the comparison where our model jointly processes the spatial and temporal dimensions of the contextual data to extract features for anomaly detection.
- 28 July 2020
- Published by MDPI AG in Applied Sciences
- Vol. 10 (15) , 5191
Open AccessA Hybrid Deep Learning Framework for Unsupervised Anomaly Detection in Multivariate Spatio-Temporal Data
Multivariate time-series data with a contextual spatial attribute have extensive use for finding anomalous patterns in a wide variety of application domains such as earth science, hurricane tracking, fraud, and disease outbreak detection. In most settings, spatial context is often expressed in terms of ZIP code or region coordinates such as latitude and longitude. However, traditional anomaly detection techniques cannot handle more than one contextual attribute in a unified way. In this paper, a new hybrid approach based on deep learning is proposed to solve the anomaly detection problem in multivariate spatio-temporal dataset. It works under the assumption that no prior knowledge about the dataset and anomalies are available. The architecture of the proposed hybrid framework is based on an autoencoder scheme, and it is more efficient in extracting features from the spatio-temporal multivariate datasets compared to the traditional spatio-temporal anomaly detection techniques. We conducted extensive experiments using buoy data of 2005 from National Data Buoy Center and Hurricane Katrina as ground truth. Experiments demonstrate that the proposed model achieves more than 10% improvement in accuracy over the methods used in the comparison where our model jointly processes the spatial and temporal dimensions of the contextual data to extract features for anomaly detection.
- 30 March 2019
- Published by World Scientific Pub Co Pte Ltd in Modern Physics Letters B
- Vol. 33 (09)
Mathematical models for phase transitions in biogels
It has been shown that reversible and irreversible phase transitions of biogels can be represented by epidemic models. The irreversible chemical sol–gel transitions are modeled by the Susceptible-Exposed-Infected-Removed (SEIR) or Susceptible-Infected-Removed (SIR) epidemic systems whereas reversible physical gels are modeled by a modification of the Susceptible-Infected-Susceptible (SIS) system. Measured sol–gel and gel–sol transition data have been fitted to the solutions of the epidemic models, either by solving the differential equations directly (SIR and SEIR models) or by nonlinear regression (SIS model). The gel point is represented as the “critical point of sigmoid,” defined as the limit point of the locations of the extreme values of its derivatives. Then, the parameters of the sigmoidal curve representing the gelation process are used to predict the gel point and its relative position with respect to the transition point, that is, the maximum of the first derivative with respect to time. For chemical gels, the gel point is always located before the maximum of the first derivative and moves backward in time as the strength of the activation increases. For physical gels, the critical point for the sol–gel transition occurs before the maximum of the first derivative with respect to time, that is, it is located at the right of this maximum with respect to temperature. For gel–sol transitions, the critical point is close to the transition point; the critical point occurs after the maximum of the first derivative for low concentrations whereas the critical point occurs after the maximum of the first derivative for higher concentrations.
- 30 March 2019
- Published by World Scientific Pub Co Pte Ltd in Modern Physics Letters B
- Vol. 33 (09)
Mathematical models for phase transitions in biogels
It has been shown that reversible and irreversible phase transitions of biogels can be represented by epidemic models. The irreversible chemical sol–gel transitions are modeled by the Susceptible-Exposed-Infected-Removed (SEIR) or Susceptible-Infected-Removed (SIR) epidemic systems whereas reversible physical gels are modeled by a modification of the Susceptible-Infected-Susceptible (SIS) system. Measured sol–gel and gel–sol transition data have been fitted to the solutions of the epidemic models, either by solving the differential equations directly (SIR and SEIR models) or by nonlinear regression (SIS model). The gel point is represented as the “critical point of sigmoid,” defined as the limit point of the locations of the extreme values of its derivatives. Then, the parameters of the sigmoidal curve representing the gelation process are used to predict the gel point and its relative position with respect to the transition point, that is, the maximum of the first derivative with respect to time. For chemical gels, the gel point is always located before the maximum of the first derivative and moves backward in time as the strength of the activation increases. For physical gels, the critical point for the sol–gel transition occurs before the maximum of the first derivative with respect to time, that is, it is located at the right of this maximum with respect to temperature. For gel–sol transitions, the critical point is close to the transition point; the critical point occurs after the maximum of the first derivative for low concentrations whereas the critical point occurs after the maximum of the first derivative for higher concentrations.
- 1 December 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Sparse Channel Estimation for Space-Time Block Coded OFDM-Based Underwater Acoustic Channels
Communication over acoustic signals underwater results in multi-scale multi-lag channels due to multipath propagation. Hence, a robust channel estimation technique has to be present at the receiver. In this paper, assuming underwater channels undergoing Rayleigh fading, a path-based channel model that characterizes each path of the time-varying sparse channel by a delay, a Doppler scale, and an attenuation factor is considered. Alamouti's space-time block transmit diversity scheme is used in the form of two transmit antennas and one receiver, and the proposed OFDM-based non-data-aided algorithm iteratively estimates the complex channel parameters of each subcarrier using the expectation maximization (EM) method, which in turn converges to a true maximum a posteriori probability (MAP) estimate of the unknown channel, where the Karhunen-Lo'eve expansion is performed for complexity reduction. Finally, the novel channel estimation algorithm combines the aforementioned MAP-EM technique with ESPRIT for delay estimation by exploiting the sparseness of the underwater acoustic channels. The performance of the proposed algorithm is then presented in terms of average mean square error and symbol error rate for QPSK signaling with extreme Doppler spreads and different pilot spacings. It is shown that excellent mean-square error and symbol error rate performance is achieved even in the presence of extreme Doppler shifts.
- 1 December 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Sparse Channel Estimation for Space-Time Block Coded OFDM-Based Underwater Acoustic Channels
Communication over acoustic signals underwater results in multi-scale multi-lag channels due to multipath propagation. Hence, a robust channel estimation technique has to be present at the receiver. In this paper, assuming underwater channels undergoing Rayleigh fading, a path-based channel model that characterizes each path of the time-varying sparse channel by a delay, a Doppler scale, and an attenuation factor is considered. Alamouti's space-time block transmit diversity scheme is used in the form of two transmit antennas and one receiver, and the proposed OFDM-based non-data-aided algorithm iteratively estimates the complex channel parameters of each subcarrier using the expectation maximization (EM) method, which in turn converges to a true maximum a posteriori probability (MAP) estimate of the unknown channel, where the Karhunen-Lo'eve expansion is performed for complexity reduction. Finally, the novel channel estimation algorithm combines the aforementioned MAP-EM technique with ESPRIT for delay estimation by exploiting the sparseness of the underwater acoustic channels. The performance of the proposed algorithm is then presented in terms of average mean square error and symbol error rate for QPSK signaling with extreme Doppler spreads and different pilot spacings. It is shown that excellent mean-square error and symbol error rate performance is achieved even in the presence of extreme Doppler shifts.
- 1 November 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-4
Orak Hücreli Anemi Tespiti Sickle Cell Anemia Detection
Anemia is a common name given to falls in oxygen transport capacity due to some of the functional disadvantages of red blood cells. Pathology Laboratorians put the tissue on the microscope glass and try to diagnose Anemia disease. Processes have been taken for a long time and it has been caused to distract. Therefore it has been caused to misdiagnose the Laboratorian. This work shortens the diagnostic period of the disease and to minimizes error probability of this diagnosis by extracting healthy cells and just having sickle cells on the blood tissue using Image Processing Algorithms with an accuracy of 91.11 %, precision of 92.9 %, recall of 79.05 % for Sickle Cell Anemia.
- 1 November 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-4
Orak Hücreli Anemi Tespiti Sickle Cell Anemia Detection
Anemia is a common name given to falls in oxygen transport capacity due to some of the functional disadvantages of red blood cells. Pathology Laboratorians put the tissue on the microscope glass and try to diagnose Anemia disease. Processes have been taken for a long time and it has been caused to distract. Therefore it has been caused to misdiagnose the Laboratorian. This work shortens the diagnostic period of the disease and to minimizes error probability of this diagnosis by extracting healthy cells and just having sickle cells on the blood tissue using Image Processing Algorithms with an accuracy of 91.11 %, precision of 92.9 %, recall of 79.05 % for Sickle Cell Anemia.
- Open Access
Channel Estimation for Visible Light Communications Using Neural Networks
Visible light communications (VLC) is an emerging field in technology and research. Estimating the channel taps is a major requirement for designing reliable communication systems. Due to the nonlinear characteristics of the VLC channel those parameters cannot be derived easily. They can be calculated by means of software simulation. In this work, a novel methodology is proposed for the prediction of channel parameters using neural networks. Measurements conducted in a controlled experimental setup are used to train neural networks for channel tap prediction. Our experiment results indicate that neural networks can be effectively trained to predict channel taps under different environmental conditions.
Open AccessChannel Estimation for Visible Light Communications Using Neural Networks
Visible light communications (VLC) is an emerging field in technology and research. Estimating the channel taps is a major requirement for designing reliable communication systems. Due to the nonlinear characteristics of the VLC channel those parameters cannot be derived easily. They can be calculated by means of software simulation. In this work, a novel methodology is proposed for the prediction of channel parameters using neural networks. Measurements conducted in a controlled experimental setup are used to train neural networks for channel tap prediction. Our experiment results indicate that neural networks can be effectively trained to predict channel taps under different environmental conditions.
- 1 May 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-4
Anomaly detection in walking trajectory
Analysis of the walking trajectory and the detection of anomalies in this trajectory, provide important benefits in the fields of health and security. In this work, two methods to detect anomalies in trajectories, are compared. Firstly, an unsupervised method is used where the conformance among trajectories are taken into consideration. Trajectories that deviate from others are qualified as anomalies. Secondly, the points in the trajectories are considered as a time series. Artifical neural networks performing supervised learning based on the backpropagation algorithm are used. The results are compared, and the points to be enhanced are highlighted.
- 1 May 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-4
Anomaly detection in walking trajectory
Analysis of the walking trajectory and the detection of anomalies in this trajectory, provide important benefits in the fields of health and security. In this work, two methods to detect anomalies in trajectories, are compared. Firstly, an unsupervised method is used where the conformance among trajectories are taken into consideration. Trajectories that deviate from others are qualified as anomalies. Secondly, the points in the trajectories are considered as a time series. Artifical neural networks performing supervised learning based on the backpropagation algorithm are used. The results are compared, and the points to be enhanced are highlighted.