AMAsparuh Markovski

Asparuh Markovski

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Asparuh Markovski has a total of 20 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Direct Neural Dynamic Programming Adaptive Control of Knee and Ankle Joints in a Simulated Prosthetic Leg Model

    by Jordan K. Kralev, Tsonyo N. Slavov, Asparuh Markovski, Vladislav Slavov

    Effective control of prosthetic legs is essential for restoring natural mobility and enhancing the quality of life for amputees. This paper presents a Direct Neural Dynamic Programming (DNDP) algorithm that leverages coupled action and critic neural networks to solve the optimal control problem for prosthetic gait trajectory tracking. By continuously learning and adapting to individual gait patterns, the DNDP action networks generate the control for the knee and ankle joints to reduce the gait tracking error. The proposed approach enables real-time optimization of control strategies without relying on precise system models, offering adaptability to subject-specific dynamics and varying conditions.

  • Open Access

    Machine Learning with Insufficient Data for Classification of Mixtures of Sunflower and Olive Oil Samples Using Laser-Induced Fluorescence Spectroscopy

    by Asparuh Markovski, Lidia Zaharieva, Vera Deneva, Elena Taskova, Tsanislava Genova, Alexander Gegov, Christina Andreeva, Liudmil Antonov

    The question of verification of food quality has stood before scientists since ancient times, and, nowadays, the advances in science and technology have made it a very challenging task. Laser-induced fluorescence (LIF) spectroscopy has become a very useful instrument for sample characterization. Nevertheless, analysis of complex multi-component spectra is difficult to approach. In recent years, the capabilities of artificial intelligence have attracted a lot of attention, as they open doors to efficient solutions of many problems that otherwise require a lot of time, effort, expenses and often inspiration. In the present work, we use LIF spectra of mixtures of sunflower and extra virgin olive oils with different concentrations and apply neural network (NN) algorithms with the aim of improving the strategies for concentration determination. Two different approaches have been applied and their output has been compared and commented. More specifically, the task of concentration recognition has been targeted as a classification and as a fitting problem. We formulate four diagnostic parameters with biochemical meaning and compare the NN performance when training with raw spectra and with the diagnostic parameters. The correct choice of appropriate diagnostic parameters is of importance from the point of view of biochemical interpretability and analysis, whereas “black box” full-spectra training might be beneficial for end-user applications. Our results show that these methods perform well even with very scarce data and outline preliminary strategies for defining diagnostic criteria.


  • Open Access

    Deep Learning for Sustainable Product Design: Shuffle-GhostNet Optimized by Enhanced Hippopotamus Optimizer to Life Cycle Assessment Integration

    by Anastasiia Rozhok, Tasho Tashev, Asparuh Markovski, Mihail Tuchin, Liubov Karnaukhova, Mikhail Ivanov

    The intelligence of sustainable design is reflected in the demands for accurate and real-time environmental impact assessments; traditional LCA methods are slow and static. In this paper, we propose a novel deep learning framework that serially links Shuffle-GhostNet (a lightweight convolutional neural network employing a combination of Ghost and Shuffle modules) improved by an enhanced version of Hippopotamus Optimizer (EHHO) for hyperparameter tuning and enhanced convergence. Upon testing the model on the Ecoinvent and OpenLCA Nexus datasets, pronounced advantages in predicting CO2 emissions, energy use, and other sustainability indicators were found. Coupling the integration of multi-source sensor data and optimizing the architecture via metaheuristic search enables rapid and reliable decision support on eco-design. Final results are significantly better than the baseline models, achieving an R2 of up to 0.943 with actual performance gains. AI-driven modeling integrated with LCA constitutes a pathway toward dynamic and scalable sustainability assessment in Industry 4.0 and circular economy applications.


  • Open Access

    Computer-assisted school experiments as a method for improvement of students’ knowledge in mechanics and atomic/molecular physics in secondary school

    by Fabien Kunis, Konstantin Ilchev, Milena Stoyanova, Vesela Dimova, Tsanislava Genova-Atanasova, Asparuh Markovski, Christina Andreeva

    We investigate the possibility to improve the effectiveness of school teaching in physics by means of introducing a partially computer-based learning experiment. More specifically, we focused on the preparation and implementation of laboratory exercises in the fields of "Mechanics" and "Atomic/molecular physics" according to the educational content in physics in secondary schools in Bulgaria. The laboratory practicum which we offer for implementation was performed by a small group of 10-th grade (16-year old) students and we present the results as a case study. The pre- and post-practice survey identified areas where significant improvement was observed, as well as some that require additional approaches to overcome difficulties. The results of the research confirm the formulated hypothesis and give reason to claim that the use of a physical experiment in physics education leads to improving the understanding of the basic concepts in the field of mechanics and atomic/molecular physics; improving skills for working with certain devices and computer programs; creating a better understanding of scientific experiment and its relationship to the physical theory.