Browse Items (16481 total)
Sort by:
-
Performance Evaluation of Convolutional Neural Networks for Stellar Image Classification: A Comparative Study
This study analyzes three distinct convolutional neural network (CNN) models, ResNet, Parallel CNN, and VGG16, for object classification using the Star-Galaxy Classification dataset. The dataset comprises a vast collection of celestial object images, including galaxies, stars, and quasars. The effectiveness of each CNN model is evaluated based on accuracy, a commonly used performance metric. The results reveal that the Parallel CNN model achieved the highest accuracy of 90.08% in classifying celestial objects, followed by VGG16 with an accuracy of 86%, and ResNet with an accuracy of 83%. Specifically, the Parallel CNN model demonstrates superior performance in classifying galaxies and stars. These findings provide valuable insights into the strengths and weaknesses of each model for this specific classification task, guiding the development of more effective CNN models for similar applications in cosmology and other fields. This research contributes to the growing literature on CNN models' application in astronomy and underscores the importance of selecting appropriate models to achieve high accuracy in object classification tasks. The study's insights can be utilized to inform the development of more effective CNN models for similar tasks and facilitate advancements in astronomical research. 2023 IEEE. -
Performance Evaluation of CPU and GPU Processors Using Advanced Data Analysis Techniques
The modern industry is developing advanced CPU and GPU processors. The standard efficiency difference between the Intel Core i5 and 11th Gen with Nvidia GTX3050 processors is being discussed in this article. However, the reduction factor between these processors is determined to be as 2.5. Various data visualization techniques were applied to give a comprehensive analysis of the performance of CPU and GPU-based processors for execution of intensive tasks. The results were analyzed to understand the various performance parameters related to their functioning and efficiency. A model was proposed for enhancing the performance and throughput of the processor by easing the internal communication process between the CPU and the GPU by converting from electrical signals to light signals, though being faced with many challenges in the current time, holds a large scope of further research in the pursuit of higher computational efficiency. The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025. -
Performance evaluation of diesel engine using genetic algorithm
?Abstract: Engine analysis and optimization is not a new approach to the field of automobiles. It has always been a keen focus in the research of experts domestically as well as internationally, the control of Air-Fuel Ratio (AFR) in transient operating conditions of engine. For the last few decades, the industry and economic expansion of developed countries has showed a clean increase in the vehicle production as well as transport volume. Global warming, acid rain, greenhouse effect and air pollution problems related to emission of CO2, NOx, PM, CO and unburned HC, together with the consumption of fossil fuels, unite to create serious problems at a global level. Therefore it is a research study considering all these current issues and taking it to a new level of optimization for the output of a better efficiency, better economy and less pollution. Performance of Diesel Engine is evaluated by parameters like Power, Torque and Specific Fuel Consumption. 2018, Blue Eyes Intelligence Engineering and Sciences Publication. All rights reserved. -
Performance Evaluation of Friction Stir Spot Welding of Al 5754 and Al 6111 using Machine Learning Approaches
This study evaluates advanced machine learning (ML) and deep learning (DL) models for predicting the tensile shear and bending strength of friction stir spot welding joints involving Al 5754 and Al 6111 alloys. ML techniques include Linear Regression, Decision Tree, Random Forest (RF), K-Nearest Neighbors, Support Vector Regression, and XGBoost, while DL models comprise Recurrent Neural Network (RNN) and Backpropagation Neural Network (BPNN). The models were assessed for discrepancies between experimental and predicted results, with the best-performing model identified using R-squared (R2), Root-Mean-Square Error, Mean Square Error, and Mean Absolute Error. The data preprocessing phase included feature scaling and an 85:15 train-test split. Key input process parameters included spindle speed, dwell time, plunge depth, and tool pin profile. The results demonstrate that XGBoost yielded the highest predictive accuracy, achieving an R2 score of 99.99% for both tensile shear and bending strength, while RF offered a strong balance between accuracy and robustness. Other ML models struggled with the datasets complexity, resulting in lower performance. Among DL approaches, the BPNN outperformed the RNN, achieving approximately 99.8% accuracy by effectively capturing complex data patterns. ASM International 2025. -
Performance Evaluation of Hybrid Lifi-Wifi Internet Systems
The rapid development of wireless technology made Light Fidelity, or LiFi, leave the laboratory and move onto the list of the next big thing, alongside everyday WiFi. WiFi also provides freedom of movement, but once everyone in a large area is streaming video, the signal becomes like a traffic jam. LiFi is going another way: data is sent via the rays of LED lights, meaning it can push files at light speed. The only problem is that the connection will not last forever, as soon as you leave the arc of the lamp or walk too far. In this paper, we explore what would happen when you combine the wide coverage of WiFi with the high speed of Li-Fi to form a hybrid system that avoids the weaknesses of both. Various test-based classrooms, halls, and open laboratories measured the speed of bit movement, the consistency of the beams, and the responsiveness of the games, and the mixed environment performed better in nearly all tests than the technology alone. It even figured out which to use as people entered and left the building, alternating between LiFi and WiFi, increasing uptime and doubling peak throughput. Hospitals, campuses, airports, and factories in need of constant, high-speed connections now have a better roadmap for using this blend so machines keep running smoothly and users remain engaged. The paper itself contains the unraveling of the hardware, the explanation of the test setup, and the weighing of the numbers, accompanied by some thought on what this combination can offer the future of wireless worlds. The findings justify integrating LiFi with WiFi as an innovative, scalable solution to ensure it keeps up with the massive data volumes generated by modern phones, tablets, and smart devices. 2025, Innovative Information Science and Technology Research Group. All rights reserved. -
PERFORMANCE EVALUATION OF IPE AND IE-AFFECTED PATIENTS USING A MODIFIED PSO AND ANFIS
Epilepsy, a complex neurological disorder, is particularly challenging to diagnose and manage when driven by genetic factors. This study focuses on the analysis of Idiopathic Partial Epilepsy (IPE) and Idiopathic Epilepsy (IE) in both children and women, using a novel approach combining Modified Particle Swarm Optimization (MPSO) with a 9-rule Adaptive Neuro-Fuzzy Inference System (ANFIS). Four feature extraction techniquesDiscrete Wavelet Transform (DWT), Shearlet Transform (SLT), Contourlet Transform (CLT), and Stockwell Transform (SWT)are employed to process electroencephalogram (EEG) signals. The performance of the proposed MPSO-ANFIS model is evaluated and compared with existing methods. Results indicate that the SWT-ANFIS-MPSO method achieves superior classification accuracy for both IE and IPE patients, highlighting its potential to improve epilepsy diagnosis and treatment strategies. 2025, Institute of Mechanics of Continua and Mathematical Sciences. All rights reserved. -
Performance Evaluation of Machine Learning Models for Detecting Vulnerabilities in Internet of Things Network
Security threats and attacks are a growing concern in the field of Internet of Things (IoT) infrastructure. Internet-based automated network application models are used across various domains; commensurately, different security vulnerabilities and anomaly attacks are also increased at the same level. These attacks could cause failures in IoT infrastructure and network systems. In the modern world, Machine Learning (ML) models support various predictive analyses, providing more accurate results for future forecasting in various fields. In this article, we compare existing classical Machine Learning (ML) algorithms supported by Artificial Intelligence (AI) to evaluate and predict the performance and accuracy of different vulnerabilities in IoT infrastructure. We considered and compared the results of Logistic Regression (LR), Support Vector Machine (SVM), Decision Tree (DT), Random Forest (RF), and Artificial Neural Network (ANN) using publicly available datasets. Through this evaluation, we obtained an accuracy of 99.4% from DT, RF, and ANN. Additionally, RF demonstrated a highest accuracy of F1 is 0.994 and lowest STD variance is 0.014 than compared models in the selected dataset. 2026, Dr. Yashwant Research Labs Pvt. Ltd. All rights reserved. -
Performance evaluation of machinelearning techniques indiabetes prediction
Diabetes diagnosis is very important at preliminary stage rather than treatment. In todays world devices like sensors are used for detection of diabetes. Accurate classification techniques are required for automatic identification of diabetes disease. In regards to research diabetes prediction with minimal number of attributes (test parameters) is to be identified earlier research states about feature reduction but with less predictive accuracy. In this regards, this work exploits machine learning techniques(methodology) such as Logistic Regression (LR), Artificial Neural Network (ANN), Support Vector Machine (SVM), Random Forest (RF) and Neural Network (NN) with 10-fold Cross Validation (CV) for classification and prediction of diabetes with Feature Selection Methods (FSMs) using R platform. Above all models enable us to investigate the relationship between a categorical outcome and a set of explanatory variables. The experiment was conducted on PIMA Indian diabetes dataset selected from UCI machine learning repository. From the experimental results it is identified that for full set of diabetes dataset attributes, Classification Accuracy (CA) achieved was 84.25%whereas with reduced set attributes an accuracy of 85.24% is achieved using NN with 10-fold CV technique compared to others which will help in medical application to predict diabetes with minimal features. BEIESP. -
Performance evaluation of Map-reduce jar pig hive and spark with machine learning using big data
Big data is the biggest challenges as we need huge processing power system and good algorithms to make a decision. We need Hadoop environment with pig hive, machine learning and hadoopecosystem components. The data comes from industries. Many devices around us and sensor, and from social media sites. According to McKinsey There will be a shortage of 15000000 big data professionals by the end of 2020. There are lots of technologies to solve the problem of big data Storage and processing. Such technologies are Apache Hadoop, Apache Spark, Apache Kafka, and many more. Here we analyse the processing speed for the 4GB data on cloudx lab with Hadoop mapreduce with varing mappers and reducers and with pig script and Hive querries and spark environment along with machine learning technology and from the results we can say that machine learning with Hadoop will enhance the processing performance along with with spark, and also we can say that spark is better than Hadoop mapreduce pig and hive, spark with hive and machine learning will be the best performance enhanced compared with pig and hive, Hadoop mapreduce jar. Copyright 2020 Institute of Advanced Engineering and Science. All rights reserved. -
Performance evaluation of multi band disk-shaped terahertz MIMO antenna with hexagon slots on ground for future 6?G and terahertz communication system
In this paper, a four-port wideband MIMO antenna is developed for future 6 G wireless communication systems. The proposed disk-shaped antenna consists of four identical disk-shaped elements arranged uniformly around a circular structure. This uniform arrangement helps maintain geometric balance and minimizes mutual coupling. All the disk patterns are arranged equidistantly around the centrally etched flower-like structure, which ensures the symmetrical geometry of the proposed antenna. These slots are helpful in generating a super-wide bandwidth ranging from 1.81 THz to 4.1513 THz. To further enhance the efficiency of the antenna, hexagonal slots are etched on the ground plane. The hexagonally etched slots on the ground reduce signal reflection losses. The overall dimensions of the four-port MIMO antenna are 800 800 50 m , and it is designed on a silicon substrate with a relative permittivity of 11.9. The proposed antenna achieves a super-wide bandwidth ranging from 0.926 THz to 5.5411 THz and a peak gain of 7 dB. MIMO performance parameters such as diversity gain, Total Active Reflection Coefficient (TARC), Envelope Correlation Coefficient (ECC), and Channel Capacity Loss (CCL) are evaluated, and all lie within acceptable ranges. The disk-shaped antenna demonstrates super-wideband characteristics, high resolution, and a low reflection coefficient. The disk-shaped antenna operates at 1.8175 THz, 2.5911 THz, 3.286 THz, and 4.1513 THz, with reflection coefficients of ?28.02 dB, ?35.723 dB, ?37.11 dB, and ?32.35 dB, respectively. Considering to its compact size, wide bandwidth, and stable radiation characteristics, the proposed disk-shaped antenna is well suited for high-speed THz communication and beyond-6G wireless applications. Copyright 2026. Published by Elsevier GmbH. -
Performance Evaluation of OTFS Under Different Channel Conditions for LEO Satellite Downlink
Orthogonal Time Frequency Space (OTFS) modulation scheme is being actively pursued as a viable alternative to Orthogonal Frequency Division Multiplexing (OFDM) modulation scheme in future wireless standards due to the inherent ability of OTFS to mitigate the Doppler effects in high mobility scenarios. The inclusion of Non Terrestrial Network (NTN) in Release 17 of 3GPP (3rd generation partnership project) New Radio (NR) standard signifies the vital role of Satellite Communications to achieve coverage extension, capability augmentation and seamless global connectivity. In this context, it becomes important to study the suitability of OTFS modulation scheme with respect to satellite channel scenarios. In this paper, we consider the downlink channel scenarios defined by 3GPP NR NTN for Low Earth Orbit (LEO) satellites at sub-6 GHz and millimetre wave frequencies for evaluating the performance of OTFS modulation schemes. Simulation results using LMMSE (Linear Minimum Mean Square Error) and MRC (Maximum Ratio Combining) detection algorithms confirm that OTFS modulation is highly robust against Doppler effects and performs consistently across all channel conditions. From simulation results, it is observed that the performance of iterative MRC detection is better than LMMSE for 16QAM and 64QAM modulation schemes by achieving respective gains of around 5 dB and 10 dB for corresponding Bit Error Rate (BER) values of 0.01 and 0.1. 2023 IEEE. -
Performance evaluation of parallel genetic algorithm for brain MRI segmentation in hadoop and spark /
Indian Journal of Science and Technology, Vol.8, Issue 48, pp.1-7, ISSN: 0974-6846 (Print), 0974-5645 (Online). -
Performance Evaluation of Predicting IoT Malicious Nodes Using Machine Learning Classification Algorithms
The prediction of malicious nodes in Internet of Things (IoT) networks is crucial for enhancing network security. Malicious nodes can significantly impact network performance across various scenarios. Machine learning (ML) classification algorithms provide binary outcomes ("yes" or "no") to accurately identify these nodes. This study implements various classifier algorithms to address the problem of malicious node classification, using the SensorNetGuard dataset. The dataset, comprising 10,000 records with 21 features, was preprocessed and used to train multiple ML models, including Logistic Regression, Decision Tree, Naive Bayes, K-Nearest Neighbors (KNN), and Support Vector Machine (SVM). Performance evaluation of these models followed the ML workflow, utilizing Python libraries such as scikit-learn, Seaborn, Matplotlib, and Pandas. The results indicated that the Naive Bayes classifier outperformed others with an accuracy of 98.1%. This paper demonstrates the effectiveness of ML classifiers in detecting malicious nodes in IoT networks, providing a robust predictive model for real-time application. The SensorNetGuard dataset is available on the IEEE data port and Kaggle platform. 2024, Prof.Dr. ?skender AKKURT. All rights reserved. -
Performance evaluation of random forest with feature selection methods in prediction of diabetes
Data mining is nothing but the process of viewing data in different angle and compiling it into appropriate information. Recent improvements in the area of data mining and machine learning have empowered the research in biomedical field to improve the condition of general health care. Since the wrong classification may lead to poor prediction, there is a need to perform the better classification which further improves the prediction rate of the medical datasets. When medical data mining is applied on the medical datasets the important and difficult challenges are the classification and prediction. In this proposed work we evaluate the PIMA Indian Diabtes data set of UCI repository using machine learning algorithm like Random Forest along with feature selection methods such as forward selection and backward elimination based on entropy evaluation method using percentage split as test option. The experiment was conducted using R studio platform and we achieved classification accuracy of 84.1%. From results we can say that Random Forest predicts diabetes better than other techniques with less number of attributes so that one can avoid least important test for identifying diabetes. Copyright 2020 Institute of Advanced Engineering and Science. All rights reserved. -
Performance Evaluation of Refractory Bodies Fabricated from Composite Oxide Powders Beneficiated from Black Al-dross
Aluminum Oxide (Corundum, ?-Al2O3) and Magnesium-Aluminum Oxides (Spinel, MgAl2O4) are highly desired refractory materials due to their ability to withstand high-temperature service conditions without corroding and cracking. They are present in composite form in black Aluminum Dross (Al-dross), a hazardous industrial waste. 1 Kg batch of this composite powder was beneficiated from Al-dross to 98+% purity after removing the hazardous Aluminum Nitride (AlN) by aqueous treatment of Al-dross in an environment-friendly manner. The treated slurry was oven dried, ball milled to fine powder, hydraulically pressed, and sintered at 1500 C/6 h into solid cylinders (50 mm diameter 20 mm height). The structural phase analysis of the sintered product (refractory blocks) revealed a highly crystalline XRD pattern with peaks pertaining to only ?-Al2O3 and MgAl2O4. The blocks with Rockwell Hardness values of 4850 HRC, were subjected to thermal shock cycling by following the guidelines of IS 1528 (heat quench between 1000 C and air at ambient) which successfully withstood > 100 shock cycles without failure. SEM was employed to study the fracture surface in an as-sintered state and after thermal shock cycling, to reveal a fine-grained microstructure with clear grain boundaries in the as-sintered state to a glassy matrix with fine cracks at the end of the thermal shock cycle test. The potential for utilization of Al-dross for refractory applications was thus established. 2023 -
Performance evaluation of ternary blended geopolymer binders comprising of slag, fly ash and brick kiln rice husk ash
The use of industrial and agro-based precursor materials from local sources can achieve desirable properties for geopolymer binders, and thus realize the carbon-efficient sustainable materials in the construction industry. At the same time, the synergy between these precursors can be assessed using the multilevel material investigation, which has not been explored extensively. Moreover, there are limited studies on ternary geopolymer synthesized with rice husk ash from uncontrolled burning source such as brick kilns. Therefore, this study evaluates the performance of ternary blended geopolymer binders comprised of ground granulated blast furnace slag (GGBFS), fly ash (FLA), and brick kiln rice husk ask (BRHA), implementing the multilevel material approach. The experimental program includes assessment and comparative analyses of the properties of geopolymer binders such as setting time, flow, compressive strength, density, water absorption, and efflorescence. Additionally, X-ray diffraction (XRD) and scanning electron microscopy (SEM) analyses examine crystallographic structure and microscopic morphology of the composite binders. The initial setting time ranged from 21 min to 47 min for ternary mixes, in comparison to 21 min to 58 min for binary mixes. GGBFS significantly contribute in setting of binder due to hydration reaction and formation of C-S-H gel. The flow of ternary mixes exhibits standard deviation of 11.42 mm when compared to 20.96 mm of binary mixes. Lower dispersion in flow values suggests improved coaction between GGBFS, FLA, and BRHA. The compressive strength of ternary mixes improved when compared to the binary mixes. The optimum performance of 60 MPa was obtained for G60A40F95R5, which was 25% and 66.67% higher than binary mixes G60F40 and G60R40, respectively. Similarly, ternary mix G70A30F95R5 showed the least water absorption of 2.08% which was 53% and 58.4% lower than the binary mixes G70F30 and G70R30, respectively. The improvement in the properties of ternary mixes was confirmed from XRD analysis, which reveal coexistence of C-S-H along with crystalline SiO2 that positively improve the microstructure of the composite binder. Moreover, SEM analysis showed dense microstructure for ternary mixes when compared to binary mixes, which further validate the improvement in the strength of such binders. The sustainability analysis discloses the enhanced performance of ternary mixes, wherein, G60A40F95R5 showed 19.35% and 46.23% lower carbon dioxide parameter than binary mixes G60F40 and G60R40, respectively. All in all, the multilevel material investigation provides a great avenue to delve in to the best performing ternary mixes which will find desirable applications in construction industry. 2024 The Authors -
Performance Evaluation of Time-based Recommendation System in Collaborative Filtering Technique
The Collaborative Filtering (CF) technique is the most common neighbourhood-based recommendation strategy, that provides personalized recommendation to a user for the items using a similarity measure. Hence, the selection of the appropriate similarity measure becomes crucial in the CF based recommendation system. The traditional similarity measures merely focus only on the historical ratings provided by the users to compute the similarity, completely ignoring the fact that preferences change over a period of time. Considering this, the paper aims to develop an effective Recommendation System that uses temporal information to capture the changes in the preferences over a period of time. For this, the existing exponential and power time decay functions are integrated with Cosine, Pearson Correlation, and Gower's similarity measures to compute similarity. The similarity is computed at the similarity computation and prediction levels of recommendation processes. Experimental findings in terms of MAE and RMSE on the MovieLens-100k demonstrate that performance of Gower's coefficient is better when applied with the exponential function at the similarity computation level of the recommendation process. 2022 Elsevier B.V.. All rights reserved. -
Performance Evaluation of Transfer Learning VGG16 in Handwritten Text Using Word Beam Search and Language Model
This study evaluates the performance of transfer learning using the VGG16 model for handwritten text recognition, integrating Word Beam Search decoding and language modeling techniques. The VGG16 model, pre-trained on large-scale datasets, serves as a feature extractor for handwritten text images, capturing intricate patterns and structures inherent in handwriting. To convert these visual features into textual information, the system employs a Recurrent Neural Network (RNN) trained with the Connectionist Temporal Classification (CTC) loss function, producing a matrix of character probabilities for each time-step. The Word Beam Search algorithm is utilized for decoding these probabilities into coherent text, effectively constructing recognized text by referencing a predefined dictionary and addressing challenges such as arbitrary character strings and varying handwriting styles. The integration of language models incorporates context which further sharpens the output and improves precision and trustworthiness of recognition systems. Experimental results demonstrate that this combined approach significantly improves recognition performance, highlighting the efficacy of transfer learning and advanced decoding strategies in handwritten text recognition. This involves analyzing its effectiveness across various datasets. Transfer learning leverages pre-trained models, like VGG16, to address challenges such as limited labeled data and extensive training times. 2025, Innovative Information Science and Technology Research Group. All rights reserved. -
Performance Improvement in E-Gun Deposited SiOx- Based RRAM Device by Switching Material Thickness Reduction
A performance improvement by reduction in switching material thickness in a e-gun deposited SiOx based resistive switching memory device was investigated. Reduction in thickness cause thinner filamentary path formation during ON-state by controlling the vacancydefects. Thinner filament cause lowering of operation current from 500 ?A to 100 ?A and also improves the reset current (from >400 ?A to <100 ?A). Switching material thickness reductionalso cause the forming free ability in the device. All these electrical parametric improvements enhance the device reliability performances. The device show >200 dc endurance, >3-hour dataretention and >1000 P/E endurance with 100 ns pulses. 2022 Institute of Physics Publishing. All rights reserved. -
Performance improvement of triple band truncated spiked triangular patch antenna
In this paper, the design of a novel triple band triangular microstrip patch antenna with inset feed is proposed. The triangular patch is designed for a resonant frequency of 2 GHz. The inset feed is placed at a depth of 1/3rd of height from the bottom of the patch for improved return loss. The insertion of two slots and two tabs causes the antenna to resonate at multiple frequencies. The proposed antenna resonates at three frequencies: 1.939 GHz, 2.515 GHz and 3.212 GHz. The truncation of the edges of the patch and the tabs improves the gain and directivity of the antenna. 2016 IEEE.