This study presents a comparative evaluation of traditional machine learning algorithms and state-of-the-art large language models (LLMs) in classification tasks using structured data. The analysis is conducted on the widely used Iris dataset, which consists of measurements from three different flower species and serves as a benchmark in machine learning literature. The performance of classical algorithms, including logistic regression, K-nearest neighbors, decision trees, and random forest, is examined based on well-established evaluation metrics. Following the implementation of these algorithms in Python (Python Software Foundation, Wilmington, DE, USA), their results are compared with the numerical reasoning capabilities of LLMs, including GPT-5.3 Instant (OpenAI, San Francisco, California, USA), Gemini 3.0 Flash (Google LLC, Mountain View, California, USA), Copilot (Microsoft Corporation, Redmond, Washington, USA), and Qwen 3.5 (Alibaba Group, Hangzhou, China). The LLMbased evaluations are performed by providing algorithm-specific prompts and analyzing the generated outputs. The findings indicate that, beyond their natural language processing capabilities, LLMs can achieve competitive performance in structured data classification tasks, reaching accuracy levels between 90% and 97%. Among the evaluated models, Gemini 3.0 Flash demonstrates superior performance by more effectively capturing multivariate relationships within the dataset. Overall, the results suggest that transformer-based architectures can serve as strong alternatives to traditional deterministic algorithms in numerical classification problems, highlighting the evolving role of LLMs as general-purpose data analysis tools.
A handheld frontend probe having a comparatively compact measurement volume has been built for the laser Doppler anemometer (LDA) system. When compared to more traditional systems, this offers substantial advantages in terms of affordability, portability, resilience, and ease of use. The probe head is unique in that it uses multi-mode fiber to gather scattered signals from the flow route and single-mode optical fiber to send the beam into the probe volume. A comparison between traditional LDA comprising a laser and a photomultiplier tube, gas laser in conjunction with an avalanche photo diode, and a portable probe head is presented in this article. The use of the newly developed system in process system measurements has a great deal of promise in the form of statistical convergence. The design and constructional factors that underlie the portable probe head are presented in this work. Its operation is then verified using data obtained from experiments through measurements of fluctuations in axial velocity in turbulent and fully developed laminar flows. To confirm that the system functions satisfactorily, these data are contrasted with theoretical distributions.
This study presents a low-profile circular microstrip patch antenna with an inset feed configuration and an H-shaped defected ground structure (DGS) for candidate frequency range three (FR3) wireless applications. The antenna was fabricated on a 1.575-mm-thick Roger RT/Duroid 5880 substrate and optimized to operate at approximately 9.5 GHz. A comparative study was conducted between a conventional full-ground antenna and the proposed H-shaped DGS configuration to investigate the effect of the DGS on antenna performance. The results show that the H-shaped DGS modifies the surface current distribution and introduces additional inductivecapacitive effects, leading to improved impedance matching and radiation characteristics. The proposed antenna shifts the maximum radiation efficiency toward the target operating frequency compared with the conventional full-ground structure, increasing the simulated radiation efficiency from 24.97% to 93.5% at 9.46 GHz. In addition, the realized gain was improved from 3.10 to 7.32 dBi while maintaining the directional radiation characteristics. Surface current analysis and an equivalent inductance-capacitance (LC) circuit model are employed to explain the physical mechanism responsible for the observed performance enhancement. The simulation results show a resonance frequency of 9.46 GHz with a reflection coefficient of −25.7 dB, a bandwidth of 400 MHz, a realized gain of 7.32 dBi, and a total efficiency of approximately 93.5%. The experimental validation demonstrates good agreement with the simulated results. The fabricated prototype achieved a measured resonance frequency of 9.50 GHz with a reflection coefficient of −30.9 dB, a bandwidth of 450 MHz, a measured gain of 6.40 dBi, and a measured radiation efficiency of approximately 77.1%. The results confirm that the proposed antenna provides a satisfactory balance between compact size, radiation performance, and fabrication simplicity, making it a promising candidate for future upper mid-band and FR3 wireless communication systems.
Neutral-to-earth voltage (NTEV) is a critical indicator of grounding integrity, insulation degradation, and electrical safety in distribution networks; however, it is often monitored only through centralized cloud-based systems, resulting in high latency, limited visibility of time-dependent anomalies, and dependency on reliable communication infrastructure. This paper presents the design, implementation, and field validation of a Smart Energy Meter–Edge Gateway (SEM-EG) architecture that integrates continuous NTEV monitoring with edgebased analytics and cloud-supported supervision. The proposed system performs real-time measurement of NTEV together with voltage, current, and power-related parameters, enabling local anomaly detection and operational awareness at the field level while providing long-term data aggregation and visualization in the cloud. Long-term deployment in an operational distribution panel demonstrates that continuous edge-based monitoring detects time-dependent grounding-related faults that remain undetected by conventional measurements, and that corrective field interventions result in an approximately 80% reduction in night-time NTEV levels, quantitatively verifying maintenance effectiveness. In addition, a multi-parameter analysis framework combining NTEV, phase currents, phase voltages, and power metrics is used to assess indirect system impacts. The SEM-EG achieves a measurement precision of ±0.5% and reduces fault detection latency by 78% compared to cloud-only approaches. Results show that edge intelligence transforms NTEV from a passive diagnostic parameter into an active, real-time operational indicator, providing a scalable and practical solution for modern smart grids.
Many types of Liénard oscillators exist in the literature and can be constructed using various semiconductor components. Liénard oscillators can use commercial memristors; however these components are costly. Therefore, researchers often rely on memristor emulators for testing. Some emulators are based on optoelectronics components that use light-sensitive components. In this study, the authors developed a low-cost memristor emulator with simple, off-the-shelf components. Two Light Emitting Diodes (LEDs) and a light-dependent resistor (LDR) were incorporated into a custom optocoupler box. A key innovation was the addition of an extra LED to bias the LDR, ensuring stable resistance and a smooth memristance. The emulator was then used to design a Liénard oscillator, demonstrating that the circuit exhibits true memristive system behavior. Mathematical models and Simulink simulations verified that the emulator introduces the required nonlinearity essential for proper oscillator operation. The simulations also revealed the expected periodic waveforms and nonlinear XY characteristics at lower operating frequencies, which are key features of Liénard systems. Finally, the circuit was tested experimentally at different frequencies. The results showed that the memristor exhibits a strong memory effect, visible in hysteresis curves, at low frequencies. At high frequencies, this effect gradually disappears. By examining the voltage–current behavior and resulting oscillation patterns, the authors verified that the circuit functions effectively as a Liénard-type oscillator.
