Research History
Ehsan's research history showcases a diverse and interdisciplinary body of work across several scientific and engineering fields. Early contributions focus on neural networks and control systems, as seen in conference publications addressing algorithms for back-propagation neural networks and learning rate optimization. Over time, the research expanded into biomedical signal processing, with notable work on arrhythmia detection based on electrocardiogram (ECG) signals, introducing a combination of morphological and time-frequency features. Ehsan's later work delves into machine learning applications for chemical and medical data. A significant portion of the research explores novel machine learning techniques, such as fuzzy wavelet networks, genetic algorithms, and adaptive neurofuzzy inference systems, applied to chemical structure analysis and QSRR modeling. The exploration of local binary patterns for ECG classification has also led to significant advancements in medical signal processing. Across this work, Ehsan has consistently applied advanced computational techniques to solve real-world problems in diverse areas, from electrical fault location to biomedical engineering, demonstrating expertise in both theoretical and applied research.
Here are some of my published papers:
- Gholamian, M., Yazdi, M., Joursaraei, A. E. Zeraatkar. An ECG classification based on modified local binary patterns: a novel approach. Res. Biomed. Eng. 37, 617–630 (2021). https://doi.org/10.1007/s42600-021-00165-0
- H. Torkaman, E. Zeraatkar, N. Deyhimi, H. H. Alhelou, and P. Siano, "Rearrangement Method of Reducing Fault Location Error in Tied Uncompleted Parallel Lines," in IEEE Access, vol. 10, pp. 51862-51872, 2022, DOI: 10.1109/ACCESS.2022.3174117.
- Fuzzy wavelet network based on extended Kalman filter training algorithm combined with least square weight estimation as a novel machine learning system in QSRR, N. Jalili-Jahani, E. Zeraatkar, Chemometrics and Intelligent Laboratory Systems, 2020, 104191,ISSN 0169-7439.
- PLS and N-PLS based MIA-QSPR modeling of the photodegradation half-lives for polychlorinated biphenyl congeners. N. Jalili-Jahani, A. Fatehi, E. Zeraatkar, RSC Advances, 2020, 10, 33753-33761
- Pixels of chemical structures correlate to chromatographic detector responses using genetic algorithmadaptive neuro-fuzzy inference system as a novel nonlinear feature selection method. N. Jalili-Jahani, E. Zeraatkar, A. Fatehi, M. Gholamian. Chemometrics and Intelligent Laboratory Systems, Volume 202, 15 July 2020.