Seda Ermis
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Seda Ermis has a total of 2 co-authors. Here are the top co-authors by number of shared publications:
- Ali Akdagli
Mersin Universitesi
1 shared publication - Mustafa Berkan Bicer
Tarsus Üniversitesi
1 shared publication
Latest Publications
- Open Access
- 13 October 2026
- Published by SAGE Publications in Building Services Engineering Research & Technology
Interpretable analytical surrogate modeling for building energy prediction via physically informed feature reduction
Accurate prediction of building energy demand is essential for efficient HVAC system design and energy-aware building operation. Although machine learning models can achieve high predictive accuracy, their limited interpretability restricts their practical use in engineering workflows. This study proposes an interpretable analytical surrogate modeling framework that integrates data-driven learning with explicit analytical formulations for building energy prediction. A systematic feature selection strategy combining correlation analysis, mutual information, and physically informed reasoning identifies a compact subset of meaningful design variables. Machine learning benchmarking demonstrates that this reduced representation preserves predictive performance comparable to the complete feature set while substantially reducing model complexity. A Generalized Additive Model (GAM) is then employed to reveal nonlinear relationships between building design variables and annual energy demand. Based on the learned GAM response functions, explicit analytical surrogate equations are derived using piecewise quadratic and low-order polynomial approximations, enabling rapid benchmark-supported design screening. The transferability of the proposed workflow is further evaluated using an independent EnergyPlus-based ASHRAE prototype building. The results demonstrate that the same methodology can be successfully applied beyond the original UCI Energy Efficiency benchmark, providing an effective balance between predictive accuracy, engineering interpretability, and practical applicability for early-stage building design. Practical application: The proposed analytical surrogate framework provides building services engineers with an efficient decision-support tool during the conceptual design stage. By combining physics-informed feature reduction with explicit analytical equations, the framework enables rapid estimation of annual heating and cooling energy demand, visualization of design trade-offs, and benchmark-supported screening of alternative building configurations without repeatedly performing computationally expensive building energy simulations. Rather than replacing detailed simulation-based analyses, the proposed workflow serves as a computationally efficient preliminary screening approach that helps designers identify promising design regions before undertaking comprehensive building performance assessment.
- 13 October 2026
- Published by SAGE Publications in Building Services Engineering Research & Technology
Open AccessInterpretable analytical surrogate modeling for building energy prediction via physically informed feature reduction
Accurate prediction of building energy demand is essential for efficient HVAC system design and energy-aware building operation. Although machine learning models can achieve high predictive accuracy, their limited interpretability restricts their practical use in engineering workflows. This study proposes an interpretable analytical surrogate modeling framework that integrates data-driven learning with explicit analytical formulations for building energy prediction. A systematic feature selection strategy combining correlation analysis, mutual information, and physically informed reasoning identifies a compact subset of meaningful design variables. Machine learning benchmarking demonstrates that this reduced representation preserves predictive performance comparable to the complete feature set while substantially reducing model complexity. A Generalized Additive Model (GAM) is then employed to reveal nonlinear relationships between building design variables and annual energy demand. Based on the learned GAM response functions, explicit analytical surrogate equations are derived using piecewise quadratic and low-order polynomial approximations, enabling rapid benchmark-supported design screening. The transferability of the proposed workflow is further evaluated using an independent EnergyPlus-based ASHRAE prototype building. The results demonstrate that the same methodology can be successfully applied beyond the original UCI Energy Efficiency benchmark, providing an effective balance between predictive accuracy, engineering interpretability, and practical applicability for early-stage building design. Practical application: The proposed analytical surrogate framework provides building services engineers with an efficient decision-support tool during the conceptual design stage. By combining physics-informed feature reduction with explicit analytical equations, the framework enables rapid estimation of annual heating and cooling energy demand, visualization of design trade-offs, and benchmark-supported screening of alternative building configurations without repeatedly performing computationally expensive building energy simulations. Rather than replacing detailed simulation-based analyses, the proposed workflow serves as a computationally efficient preliminary screening approach that helps designers identify promising design regions before undertaking comprehensive building performance assessment.
- Open Access
- 8 June 2026
- Published by Springer Nature in Energy Informatics
- Vol. 9 (1)
Improving short-term HVAC load forecasting via temporal feature engineering: insights from the PLEIAData building dataset
- 8 June 2026
- Published by Springer Nature in Energy Informatics
- Vol. 9 (1)
Open AccessImproving short-term HVAC load forecasting via temporal feature engineering: insights from the PLEIAData building dataset
- Open Access
- 17 March 2026
- Published by Springer Nature in Applied Physics A
- Vol. 132 (4) , 297
Tunable multiband metamaterial absorber with polarization sensitivity via multi-tier strip modifications for millimeter-Wave applications
This paper presents a fully passive and geometrically reconfigurable metamaterial absorber capable of achieving multi-band operation in the 20–40 GHz millimeter-wave range. The proposed design evolves systematically from a polarization-insensitive four-band configuration to polarization-sensitive five- and six-band configurations through multi-tier strip modifications applied to a single-layer metallic pattern. Without relying on active components or multilayer stacking, the absorber enables controlled generation of additional resonances and fine frequency tuning while maintaining a simple and fabrication-friendly geometry. Numerical results demonstrate that the four-band design achieves normalized absorptivity exceeding 96%, whereas the five- and six-band configurations maintain absorptivity levels above 95% and 97%, respectively. Detailed parametric analyses reveal the influence of geometric variations on resonance behavior, and electric and magnetic field as well as surface current distributions are investigated to clarify the underlying absorption mechanisms. Owing to its passive configuration, structural simplicity, and controllable polarization response, the proposed absorber offers a practical solution for millimeter-wave applications including advanced wireless systems and frequency-selective electromagnetic devices.
- 17 March 2026
- Published by Springer Nature in Applied Physics A
- Vol. 132 (4) , 297
Open AccessTunable multiband metamaterial absorber with polarization sensitivity via multi-tier strip modifications for millimeter-Wave applications
This paper presents a fully passive and geometrically reconfigurable metamaterial absorber capable of achieving multi-band operation in the 20–40 GHz millimeter-wave range. The proposed design evolves systematically from a polarization-insensitive four-band configuration to polarization-sensitive five- and six-band configurations through multi-tier strip modifications applied to a single-layer metallic pattern. Without relying on active components or multilayer stacking, the absorber enables controlled generation of additional resonances and fine frequency tuning while maintaining a simple and fabrication-friendly geometry. Numerical results demonstrate that the four-band design achieves normalized absorptivity exceeding 96%, whereas the five- and six-band configurations maintain absorptivity levels above 95% and 97%, respectively. Detailed parametric analyses reveal the influence of geometric variations on resonance behavior, and electric and magnetic field as well as surface current distributions are investigated to clarify the underlying absorption mechanisms. Owing to its passive configuration, structural simplicity, and controllable polarization response, the proposed absorber offers a practical solution for millimeter-wave applications including advanced wireless systems and frequency-selective electromagnetic devices.
- Open Access
- 29 December 2024
- Published by Celal Bayar University Journal of Science in Celal Bayar Üniversitesi Fen Bilimleri Dergisi
- Vol. 20 (4) , 40-59
The Effect of Substrate Dielectric Constant and Thickness on Millimeter Wave Band Patch Antenna Performance
In last decade, the fifth generation of telecom network (5G) has been a new era as a result of fast-growing mobile industry. Unlike its predecessors, 5G will not only provide faster, better mobile broadband experience, but also broaden communication network with new services such as device-to-device communications or connecting IoT devices and users. For this purpose, 5G aims to achieve massive network capacity, ultra-low latency, higher data speed and greater network reliability. According to the report of ITU World Radiocommunication Conference 2019 (WRC-19), several new frequency bands between 20-70 GHz, were announced for allocation of 5G. Frequencies in the Ka-band (27-40 GHz) are particularly attractive due to their low atmospheric attenuation. At the specified frequency range i.e., millimeter wave band, antenna design for 5G applications is very crucial to provide high gain and efficiency as well as broadband communication which is indispensable for high-speed data traffic. At this point, Microstrip patch antennas, stand out amongst others because of their numerous attractive features. In this work, the effect of substrate dielectric constant and thickness to the Rectangular Microstrip Antenna (RMA) performance is examined for high frequency 5G applications. The RMA is designed to operate at 38 GHz and antenna performance has been analyzed according to various dielectric substrates, such as RT5880, RO3003, FR4, RT6006 and RT6010, considering different dielectric constants, thicknesses, and tangential losses. All designs and analyses have been accomplished by using ANSYS HFSS (High-Frequency Structure Simulator) and comparative results of the work are presented.
- 29 December 2024
- Published by Celal Bayar University Journal of Science in Celal Bayar Üniversitesi Fen Bilimleri Dergisi
- Vol. 20 (4) , 40-59
Open AccessThe Effect of Substrate Dielectric Constant and Thickness on Millimeter Wave Band Patch Antenna Performance
In last decade, the fifth generation of telecom network (5G) has been a new era as a result of fast-growing mobile industry. Unlike its predecessors, 5G will not only provide faster, better mobile broadband experience, but also broaden communication network with new services such as device-to-device communications or connecting IoT devices and users. For this purpose, 5G aims to achieve massive network capacity, ultra-low latency, higher data speed and greater network reliability. According to the report of ITU World Radiocommunication Conference 2019 (WRC-19), several new frequency bands between 20-70 GHz, were announced for allocation of 5G. Frequencies in the Ka-band (27-40 GHz) are particularly attractive due to their low atmospheric attenuation. At the specified frequency range i.e., millimeter wave band, antenna design for 5G applications is very crucial to provide high gain and efficiency as well as broadband communication which is indispensable for high-speed data traffic. At this point, Microstrip patch antennas, stand out amongst others because of their numerous attractive features. In this work, the effect of substrate dielectric constant and thickness to the Rectangular Microstrip Antenna (RMA) performance is examined for high frequency 5G applications. The RMA is designed to operate at 38 GHz and antenna performance has been analyzed according to various dielectric substrates, such as RT5880, RO3003, FR4, RT6006 and RT6010, considering different dielectric constants, thicknesses, and tangential losses. All designs and analyses have been accomplished by using ANSYS HFSS (High-Frequency Structure Simulator) and comparative results of the work are presented.
- Open Access
- 1 January 2011
- Published by The Scientific and Technological Research Council of Turkey (TUBITAK-ULAKBIM) - DIGITAL COMMONS JOURNALS in Turkish Journal of Electrical Engineering and Computer Sciences
- Vol. 19 (4) , 597-606
A novel expression for resonant length obtained by using artificial bee colony algorithm in calculating resonant frequency of C-shaped compact microstrip antennas
This paper presents a novel and simple expression for resonant length to calculate the resonant frequency of C-shaped compact microstrip antennas operating on UHF band applications. C-shaped compact microstrip antennas with different physical dimensions and electrical parameters were simulated by means of a software package that employs the method of finite difference time domain. With the aid of the artificial bee colony algorithm, an expression for the resonant length depending on physical dimensions was constructed by using simulation data. The resonant length expression provided less than 1.6% error on average over the simulated 144 antennas. A comparison between the results obtained in this work and previous results presented in the literature is given to show the accuracy of the proposed expression.
- 1 January 2011
- Published by The Scientific and Technological Research Council of Turkey (TUBITAK-ULAKBIM) - DIGITAL COMMONS JOURNALS in Turkish Journal of Electrical Engineering and Computer Sciences
- Vol. 19 (4) , 597-606
Open AccessA novel expression for resonant length obtained by using artificial bee colony algorithm in calculating resonant frequency of C-shaped compact microstrip antennas
This paper presents a novel and simple expression for resonant length to calculate the resonant frequency of C-shaped compact microstrip antennas operating on UHF band applications. C-shaped compact microstrip antennas with different physical dimensions and electrical parameters were simulated by means of a software package that employs the method of finite difference time domain. With the aid of the artificial bee colony algorithm, an expression for the resonant length depending on physical dimensions was constructed by using simulation data. The resonant length expression provided less than 1.6% error on average over the simulated 144 antennas. A comparison between the results obtained in this work and previous results presented in the literature is given to show the accuracy of the proposed expression.