Performance Evaluation Of Algorithms Research Articles

Performance evaluation of image processing algorithms for eye blinking detection

The eye is considered a rich source for gathering information on our daily lives. In addition, the necessity to classify the eye as open or closed and detect eye blinks is increasing in various fields. For instance, it has been proven that eye blinks can be related to different ophthalmological conditions.Nevertheless, accurate, and robust blink estimation poses significant challenges to the efficacy of real-time applications due to the variability of eye images and light conditions.In this paper eye blinking has been monitored using a chin rest and an iOS smartphone attached to a cartesian machinery. Five video-based eye blink algorithms to classify frames have been analyzed and compared in terms of performance. To evaluate the results, different widely used statistical metrics have been employed. The results show that the proposed techniques successfully detect eye blinks, whereas the frame classifier shows different scores depending on the algorithm used.

Design of a Collaborative Vehicle Formation Control Simulation Test System

The purpose of this research is to tackle one of the most difficult issues in the realm of self-driving cars, which is the testing of advanced self-driving application scenarios. Thus, this study proposes a simulation testing system based on hardware-in-the-loop simulation technology. This system can enable data exchange between hardware systems as well as replicate and evaluate the algorithmic operations of the equipment under laboratory conditions. The system can integrate scenario simulation software with MATLAB to evaluate algorithm performance. The vehicle formation control system is tailored for collaborative vehicle formation management scenarios and tested in the simulation test system. The findings display the functional integrity of the vehicle formation control system, the reliability of lane changing and the stability and safety of cruising. It additionally demonstrates that the simulation testing system has the ability to recreate cooperative vehicle arrangement management situations and assess their functionality and performance. In forthcoming research, comprehensive functional and performance assessments will be executed on various typical scenarios for advanced autonomous driving applications in order to authenticate the simulation test system’s applicability.

Deep Learning for Logo Detection: A Survey

Logo detection has gradually become a research hotspot in the field of computer vision and multimedia for its various applications, such as social media monitoring, intelligent transportation, and video advertising recommendation. Recent advances in this area are dominated by deep learning-based solutions, where many datasets, learning strategies, network architectures, and loss functions have been employed. This article reviews the advance in applying deep learning techniques to logo detection. First, we discuss a comprehensive account of public datasets designed to facilitate performance evaluation of logo detection algorithms, which tend to be more diverse, more challenging, and more reflective of real life. Next, we perform an in-depth analysis of the existing logo detection strategies and their strengths and weaknesses of each learning strategy. Subsequently, we summarize the applications of logo detection in various fields, from intelligent transportation and brand monitoring to copyright and trademark compliance. Finally, we analyze the potential challenges and present the future directions for the development of logo detection. This study aims better to inform readers about the current state of logo detection and encourage more researchers to get involved in logo detection.

Performance evaluation of evolutionary algorithms for analog integrated circuit design optimisation

An automated sizing approach for analog circuits using evolutionary algorithms is presented in this paper. A targeted search of the search space has been implemented using a particle generation function and a repair-bounds function that has resulted in faster convergence to the optimal solution. The algorithms are tuned and modified to converge to a better optimal solution with less standard deviation for multiple runs compared to standard versions. Modified versions of the artificial bee colony optimisation algorithm, genetic algorithm, grey wolf optimisation algorithm, and particle swarm optimisation algorithm are tested and compared for the optimal sizing of two operational amplifier topologies. An extensive performance evaluation of all the modified algorithms showed that the modifications have resulted in consistent performance with improved convergence for all the algorithms. The implementation of parallel computation in the algorithms has reduced run time. Among the considered algorithms, the modified artificial bee colony optimisation algorithm gave the most optimal solution with consistent results across multiple runs.

Performance Evaluation of Machine Learning Algorithms for Sarcopenia Diagnosis in Older Adults.

Sarcopenia is a progressive and generalized skeletal muscle disorder. Early diagnosis is necessary to reduce the adverse effects and consequences of sarcopenia, which can help prevent and manage it in a timely manner. The aim of this study was to identify the important risk factors for sarcopenia diagnosis and compare the performance of machine learning (ML) algorithms in the early detection of potential sarcopenia. A cross-sectional design was employed for this study, involving 160 participants aged 65 years and over who resided in a community. ML algorithms were applied by selecting 11 features-sex, age, BMI, presence of hypertension, presence of diabetes mellitus, SARC-F score, MNA score, calf circumference (CC), gait speed, handgrip strength (HS), and mid-upper arm circumference (MUAC)-from a pool of 107 clinical variables. The results of the three best-performing algorithms were presented. The highest accuracy values were achieved by the ALL (male + female) model using LightGBM (0.931), random forest (RF; 0.927), and XGBoost (0.922) algorithms. In the female model, the support vector machine (SVM; 0.939), RF (0.923), and k-nearest neighbors (KNN; 0.917) algorithms performed the best. Regarding variable importance in the ALL model, the last HS, sex, BMI, and MUAC variables had the highest values. In the female model, these variables were HS, age, MUAC, and BMI, respectively. Machine learning algorithms have the ability to extract valuable insights from data structures, enabling accurate predictions for the early detection of sarcopenia. These predictions can assist clinicians in the context of predictive, preventive, and personalized medicine (PPPM).

Relative performance evaluation of machine learning algorithms for land use classification using multispectral moderate resolution data

Analyses of spatial and temporal patterns of land use and land cover through multi-resolution remote sensing data provide valuable insights into landscape dynamics. Land use changes leading to land degradation and deforestation have been a prime mover for changes in the climate. This necessitates accurately assessing land use dynamics using a machine-learning algorithm’s temporal remote sensing data. The current study investigates land use using the temporal Landsat data from 1973 to 2021 in Chikamagaluru district, Karnataka. The land cover analysis showed 2.77% decrease in vegetation cover. The performance of three supervised learning techniques, namely Random Forest (RF), Support Vector Machine (SVM), and Maximum Likelihood classifier (MLC) were assessed, and results reveal that RF has performed better with an overall accuracy of 90.22% and a kappa value of 0.85. Land use classification has been performed with supervised machine learning classifier Random Forest (RF), which showed a decrease in the forest cover (48.91%) with an increase of agriculture (6.13%), horticulture (43.14%) and built-up cover (2.10%). Forests have been shrinking due to anthropogenic forces, especially forest encroachment for agriculture and industrial development, resulting in forest fragmentation and habitat loss. The fragmentation analysis provided the structural change in the forest cover, where interior forest cover was lost by 27.67% from 1973 to 2021, which highlights intense anthropogenic pressure even in the core Western Ghats regions with dense forests. Temporal details of the extent and condition of land use form an information base for decision-makers.

CLASSIFICATION OF STUDENT GRADUATION STATUS USING XGBOOST ALGORITHM

College is an optional final stage in formal education. At this time, universities are required to have good quality by utilizing all the resources they have. Therefore, efforts are needed to help the faculty and study program make policies and decisions. One of the efforts that can be made is to classify student graduation status as early as possible to increase the number of students graduating on time. Thus, a classification algorithm is needed to avoid overfitting and produce good accuracy. The purpose of this study was to classified the student graduation status of the Statistics Untan Study Program using the XGBoost algorithm. XGBoost is an ensemble algorithm obtained through the development of gradient boosting. XGBoost has several features that can be used to prevent overfitting, but it can only process numerical data. Therefore, 140 numerical data were adjusted using the dummy technique in this study. The resulting XGBoost classification model is optimal at the number of rounds is 3 and the number of folds is 5. Based on the performance evaluation of the XGBoost algorithm, an accuracy of 75,00%, precision of 88,89%, sensitivity of 76,19% and specificity of 71,43% were obtained. Thus, the performance of the XGBoost algorithm is classified as good.

Performance Evaluation of Block-Sized Algorithms for Majority Vote in Facial Recognition

Facial recognition (FR) is a pattern recognition problem, in which images can be considered as a matrix of pixels.There are manychallenges that affect the performance of face recognitionincluding illumination variation, occlusion, and blurring. In this paper,a few preprocessing techniques are suggested to handle the illumination variationsproblem. Also, other phases of face recognition problems like feature extraction and classification are discussed. Preprocessing techniques like Histogram Equalization (HE), Gamma Intensity Correction (GIC), and Regional Histogram Equalization (RHE) are tested inthe AT&T database. For feature extraction, methods such as Principal Component Analysis (PCA), Linear Discriminant Analysis (LDA), Independent Component Analysis (ICA), and Local Binary Pattern (LBP) are applied. Support Vector Machine (SVM) is used as the classifier. Both holistic and block-based methods are tested using the AT&T database. For twelve different combinations of preprocessing, feature extraction, and classification methods, experiments involving various block sizes are conducted to assess the computation performance and recognition accuracy for the AT&T dataset.Using the block-based method, 100% accuracy is achieved with the combination of GIC preprocessing, LDA feature extraction,and SVM classification using 2x2 block-sizingwhile the holistic method yields the maximum accuracy of 93.5%. The block-sized algorithm performs better than the holistic approach under poor lighting conditions.SVM Radial Basis Function performs extremely well on theAT&Tdataset for both holistic and block-based approaches.

Performance evaluation and comparative analysis of CrowWhale-energy and trust aware multicast routing algorithm

Multipath routing helps to establish various quality of service parameters, which is significant in helping multimedia broadcasting in the Internet of Things (IoT). Traditional multicast routing in IoT mainly concentrates on ad hoc sensor networking environments, which are not approachable and vigorous enough for assisting multimedia applications in an IoT environment. For resolving the challenging issues of multicast routing in IoT, CrowWhale-energy and trust-aware multicast routing (CrowWhale-ETR) have been devised. In this research, the routing performance of CrowWhale-ETR is analyzed by comparing it with optimization-based routing, routing protocols, and objective functions. Here, the optimization-based algorithm, namely the Spider Monkey Optimization algorithm (SMO), Whale Optimization Algorithm (WOA), Dolphin Echolocation Optimization (DEO) algorithm, Water Wave Optimization (WWO) algorithm, Crow Search Algorithm (CSA), and, routing protocols, like Ad hoc On-Demand Distance Vector (AODV), CTrust-RPL, Energy-Harvesting-Aware Routing Algorithm (EHARA), light-weight trust-based Quality of Service (QoS) routing, and Energy-awareness Load Balancing-Faster Local Repair (ELB-FLR) and the objective functions, such as energy, distance, delay, trust, link lifetime (LLT) and EDDTL (all objectives) are utilized for comparing the performance of CrowWhale-ETR. In addition, the performance of CrowWhale-ETR is analyzed in terms of delay, detection rate, energy, Packet Delivery Ratio (PDR), and throughput, and it achieved better values of 0.539 s, 0.628, 78.42%, 0.871, and 0.759 using EDDTL as fitness.

The evaluation of convolutional neural network and genetic algorithm performance based on the number of hyperparameters for english handwritten recognition

<span>Convolutional neural network (CNN) has been widely applied to image recognition, especially Handwritten English Recognition. CNN's performance is good if the hyperparameter values are correct. However, the determination of precise hyperparameters is not a trivial task. This task is made more difficult when combined with a larger number of hyperparameters resulting in a high dimensionality of the search space. Usually, hyperparameter optimization uses a finite number. Previous studies have shown that a large number of hyperparameters can result in optimal CNN performance. However, the studies only apply to text mining datasets. This study offers two novelties. First, it applied 20 hyperparameters and their ranges to handwritten English. Second, this paper conducted seven experiments based on different hyperparameters and the number of hyperparameters. This paper also compares the existing methods, namely random and grid search. The experiment resulted in the proposed model being superior to the existing methods. EX3 is better than other experiments and a larger number of hyperparameters and layer-specific hyperparameter values are unimportant.</span>

Performance evaluation of deep learning algorithms for heat loss damage classification in buildings from UAV-borne infrared images

Deep learning (DL) algorithms are gaining popularity to automate anomaly detection in a variety of civil and mechanical engineering domains. Nevertheless, very few attempts have been made to use infrared (IR) images with DL algorithms to detect heat loss in a building and expedite the otherwise cumbersome energy audit process. However, no clear guidelines exist for the selection of appropriate algorithms or hyperparameters (e.g., number of epochs, learning rate, number of images) to implement DL with IR images. To address the research gap, a comparative sensitivity analysis of four well-known DL algorithms for the classification of heat losses using only IR images was performed for the first time. Four DL algorithms, namely, general convolutional neural network (CNN), VGG16, transfer learning (TL) VGG16, and TL inceptionV3 were selected and used for automating the detection of heat loss from IR images. The sensitivity of the selected CNN algorithms was analyzed based on i) the size of the IR images dataset, ii) the number of epochs, and iii) the effect of an unbalanced dataset. The goal was to provide a framework that may aid in the selection of suitable CNN architecture as well as parameters related to data size and an optimal number of training iterations for reliable heat loss identification using IR images. This article has provided a framework for automated heat loss detection that can be applied to any IR dataset without the need to pre-process the IR images. The accuracy of the selected DL algorithms is validated with IR images of laboratory setups and real-world buildings characterized by different types of defects. The comparative sensitivity analysis was conducted based on multiple performance indicators such as precision, recall, and F-1 score which resulted in optimal values of hyperparameters and the most suitable DL algorithm. The automated identification of heat loss yielded performance indicator values of more than 90%.

Real-Time IoV Task Offloading through Dynamic Assignment of SDN Controllers: Algorithmic Approaches and Performance Evaluation

Task offloading in Internet of Vehicles (IoV) is very crucial. The widespread use of IoT applications frequently interacts with the cloud, thereby increasing the load on centralized cloud controllers. Centralized network management in cloud infrastructure is not feasible for the latest IoT trends. Decentralized and decoupled network management in Software Defined Networks (SDN) can enhance IoV services. SDN and IoV coupling can better handle task offloading in ubiquitous and dynamic IoV environments. However, appropriate SDN controller assignment and allotment strategies play a prominent role in IoV communication. In this study, we developed algorithms for SDN controller assignment and allotment namely 1) Next Fit Allotment and Assignment of SDN Controller in IoV (NFAAC), 2) Dynamic Bin Packing Allotment and Assignment of SDN Controller in IoV (DBPAAC), and 3) Dynamic Focused and Bidding Allotment and Assignment algorithm of SDN Controller in IoV (DFBAAC). These algorithms were simulated using open-flow switch controllers. The controllers were modeled as Road Side Units (RSU) that can allocate bandwidth and resource requirements to vehicles on the road. Our results show that our proposed algorithm works efficiently for SDN controller assignment and allocation, outperforming the existing work by a significant improvement of 13.5%. The working of the proposed algorithms are verified, tested, and analytically presented in this study.

Performance Evaluation of Machine Learning Nave Bayes Algorithms for Network Traffic Classification

Network Traffic Classification (NTC) is an important field for different network statistics like management, malware detection and other paramount constraints. Artificial Intelligence (AI) including Machine Learning (ML) and Deep Learning (DL), on the other hand, plays a very important field nowadays due to its significant capabilities with an extremely different fields and for complex problems. ML, specifically, provides tools in the most important network fields like traffic management, security, etc. Performance evaluation is a very important aspect of any system. This research paper provides a method for NTC using ML Nave Bayes (NB) algorithm in terms of Bernoulli, Multinominal and Gaussian for classifying captured network traffic in two different datasets and perform a performance evaluation and a comparison among these algorithms. The first dataset is a VPN-nonVPN (ISCXVPN2016) dataset while the second is a packet-captured regular Wi-Fi traffic flow dataset for video browsing on the web. Results were comparable in terms of f1-score, accuracy and processing time. Bernoulli NB provides average 93.05% accuracy with 742 ms, Multinominal NB provides average 98.78% accuracy with 78.3 ms processing time and finally, Gaussian NB provides average 69.14% with 46.85 ms processing time.

Performance Evaluation of Some Clustering Algorithms under Different Validity Indices

Performance Evaluation of Some Clustering Algorithms under Different Validity Indices

Task‐oriented resource prediction and adjustment algorithm for space‐based information network

SummaryThe tasks of a space‐based information network are complex and diverse, but the resources of a space‐based environment are minimal. The existing methods are challenging to match task demand to resource supply accurately. Aiming at the problem of accurate prediction from task to resource, we propose a resource prediction adjustment strategy. First, we propose a multidimensional resource prediction algorithm based on improved particle swarm optimization and back propagation (IPSO‐BP) neural network. The improved PSO is used to optimize the weight and threshold of BP neural network to make up for the defects that BP neural network is easy to fall into local minimum and the predicted output value is not unique. Second, to meet the quality of service (QoS) of tasks, we propose a density‐based performance evaluation algorithm (DPEA) to adjust resources. This method uses the idea of local sensitive hash to select the evaluation subset for the configuration task, then dynamically selects the nearest neighbors of the configuration task, and uses the idea of weighted average to evaluate the QoS performance index of the configuration task. Simulation results show that the proposed resource prediction and adjustment strategy effectively reduces the scheduling time overhead and improves resource utilization.

Analysis and Performance Evaluation of Transfer Learning Algorithms for 6G Wireless Networks

The development of the 5G network and the transition to 6G has given rise to multiple challenges for ensuring high-quality and reliable network services. One of these main challenges is the emergent intelligent defined networks (IDN), designed to provide highly efficient connectivity, by merging artificial intelligence (AI) and networking concepts, to ensure distributed intelligence over the entire network. To this end, it will be necessary to develop and implement proper machine learning (ML) algorithms that take into account this new distributed nature of the network to represent increasingly dynamic, adaptable, scalable, and efficient systems. To be able to cope with more stringent service requirements, it is necessary to renew the ML approaches to make them more efficient and faster. Distributed learning (DL) approaches are shown to be effective in enabling the possibility of deploying intelligent nodes in a distributed network. Among several DL approaches, transfer learning (TL) is a valid technique to achieve the new objectives required by emerging networks. Through TL, it is possible to reuse ML models to solve new problems without having to recreate a learning model from scratch. TL, combined with distributed network scenarios, turns out to be one of the key technologies for the advent of this new era of distributed intelligence. The goal of this paper is to analyze TL performance in different networking scenarios through a proper MATLAB implementation.

Research and Development of Machine Vision Algorithm Performance Evaluation System in Complex Scenes

In recent years, deep learning technology has gained widespread attention and has been successfully applied in various applications, including machine vision for automatic driving. However, one of the major challenges in this field is the ability of AI models to perform well in complex scenes with varying weather conditions and lighting conditions. The proposed scheme uses VR technology to simulate different weather conditions and lighting conditions and to evaluate the performance of AI products in such complex scenes. To further enhance the generalization capabilities of AI models, the Software Development Kit has been updated with complex scene data with configurable parameters, and training data with typical characteristics according to machine vision tasks. This approach has the potential to significantly improve the accuracy and reliability of machine vision systems for automatic driving, ultimately enhancing their safety and effectiveness.

Modified Needleman-Wunsch algorithm for trauma management performance evaluation

BackgroundTrauma injuries are one of the leading causes of death in the world, representing approximately 8 % of all deaths. Therefore, trauma management training is of great importance and new training courses have arisen during the last decades. However, actual training courses do not typically analyze compliance with the protocols and guidelines available in the literature. Considering general trauma management guidelines such as the Advanced Trauma Life Support (ATLS) manual and the expertise of trauma specialists, a trauma management automated evaluation system has been designed in this paper. MethodsA modification to the Needleman-Wunsch (NW) algorithm is developed, including all relevant aspects of trauma management to automatically evaluate how a trauma intervention has been implemented according to trauma protocols. This allows to consider more information with respect to the order of the actions taken and the type of actions performed than current evaluation methods, such as checklists or videos recorded in simulation. A web-based trauma simulator is used so that it can be used at any setting with internet connection. Final-year medical students and first- and second-year residents performed an experimental test, where a trauma score is obtained with the modified NW algorithm. This automatic score relates to how similar the actions are to trauma protocols. ResultsThe results show the best combination of the scores used for the modified NW variables. This combination has an error, for the different case scenarios created, below 0.07 which verifies the values obtained. Additionally, trauma experts verified the results obtained showing a median difference of 0 between the protocol adherence evaluation using the algorithm and the one provided by the trauma experts. ConclusionsThe best set of score values to apply to the modified NW algorithm show that the modified NW algorithm provides a successful objective measurement with respect to the protocol compliance.

Biclustering Performance Evaluation of Cheng and Church Algorithm and Iterative Signature Algorithm

Biclustering Performance Evaluation of Cheng and Church Algorithm and Iterative Signature Algorithm

Performance Evaluation of Random Forest Algorithm for DDoS Attacks Detection in Wireless Sensor Network Using Numericalization and Normalization as Feature Engineering Techniques

Distributed Denial-Of-Service (DDoS) attacks in Wireless Sensor Networks (WSNs) aims to decrease the availability of a service by exhausting the network resources available for traffic, thus preventing legitimate users from accessing the services. This paper evaluates the performance of Random Forest algorithm for distributed denial of service attacks detection in Wireless Sensor Networks (WSNs) with numericalization and normalization as feature engineering techniques and exploratory data analysis to build the Distributed Denial of Service (DDoS) attacks detection model. The dataset was streamlined for optimum performance of the selected algorithms. The experimental result shows the Random Forest achieved an accuracy of 90%, F1 score of 90%, precision of 91% and recall of 90%. The outcomes revealed that the random forest model performed better than the competing models in terms of accuracy, F1 score, precision, and recall. The results of this study shed light on the efficiency of machine learning models in identifying DDoS attacks and suggest areas for future study. The study stresses the significance of data preparation and cleaning in the process and offers a framework for the application of machine learning models in network intrusion detection systems. Keywords: Wireless Sensor Networks, Distributed Denial of Service, Intrusion Detection System, Random Forest, Machine Learning. Proceedings Citation Format Sulaiman, H, Oyefolahan, I.O., Bashir, S.A., Kolo, I.M. & Ndunagu J.N. (2023): Performance Evaluation of Random Forest Algorithm for DDoS Attacks Detection in Wireless Sensor Network Using Numericalization and Normalization as Feature Engineering Techniques. Proceedings of the Cyber Secure Nigeria Conference. Nigerian Army Resource Centre (NARC) Abuja, Nigeria. 11-12th July, 2023. Pp 19-26. https://cybersecurenigeria.org/conference-proceedings/volume-2-2023/ dx.doi.org/10.22624/AIMS/CSEAN-SMART2023P2.