Performance Of Algorithm Research Articles

Analisis Kinerja Algoritma Machine Learning dalam Prediksi Harga Cryptocurrency

Cryptocurrency has become an increasingly popular digital asset in recent years. However, cryptocurrency prices are highly volatile and difficult to predict due to being influenced by many factors such as market sentiment, regulations, and technological adoption. This study aims to analyze the performance of several popular machine learning algorithms in accurately predicting cryptocurrency prices. We evaluated four algorithms: Linear Regression, Random Forest, Support Vector Machine (SVM), and Long Short-Term Memory (LSTM) using historical price datasets of Bitcoin, Ethereum, and Litecoin. The data were analyzed by preprocessing steps such as normalization and splitting into training and testing sets. Evaluation metrics used were Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and prediction accuracy. The experimental results showed that the LSTM algorithm had the best performance in predicting cryptocurrency prices with the highest accuracy and lowest error, followed by SVM, Random Forest, and Linear Regression. Further analysis revealed that LSTM was able to capture patterns and trends in complex time series data.

Benchmarking battery management system algorithms - Requirements, scenarios and validation for automotive applications

State estimators are crucial for the effective use of batteries in real-world applications. Insufficient algorithms can lead to user dissatisfaction, safety risks, and accelerated battery degradation, posing significant risks to manufacturers. Developing algorithms for battery management systems (BMS) involves defining requirements, implementing algorithms, and validating them, which is a complex process. The performance of BMS algorithms is influenced by constraints related to hardware, data storage, calibration processes during development and use, and costs. Additionally, state estimation methods vary widely, requiring specific data that impact algorithm performance.This study investigates these complexities in the development of state estimators and underscores the importance of their performance. We established an approach for selecting test scenarios, based on expert interviews, which considers computational capabilities and specific application scenarios. A model-based simulation environment is introduced to handle the complexities of validation. This environment enables thorough validation of the algorithms under real-application conditions, different test scenarios, and parameter variations.We exemplarily perform a validation for three State of Charge (SoC) estimators under diverse conditions and cell variations. The results show the performance dependencies on temperatures, cell chemistries, initial SoCs and measurement inaccuracies. Additionally, the cell-to-cell variations highlight the complexity and effort of algorithm validation. Introducing an additional scenario parameter expands the range of test scenarios, emphasizing the necessity to select scenarios that accurately reflect field conditions and worst-case situations.

Visual analytics of an interpretable prediction model for the glass transition temperature of fluoroelastomers

In development, precise measurement and control of Tg is essential for researchers to thoroughly understand and master how variations in formulation and production processes impact product performance. Given the complexity of industrial production, most manufacturers repeatedly produce fluorine materials and test their Tg until it satisfies the application environment’s demands. Consequently, this study utilized an FKM dataset from actual synthesis trials, subjected to data classification, alignment, and feature filtering, to develop a predictive model based on Elastic Net Regression, enhancing the accuracy of Tg predictions for FKM. We evaluated the performance of several regression machine learning algorithms, including Random Forest, Support Vector Machines, Decision Trees, Extreme Gradient Boosting, and Ensemble Learning. The findings show that Random Forest and Elastic Net models exceeded other models by approximately 5% to 10 % in the coefficient of determination, indicating superior predictive capabilities. The Elastic Net Regression model excelled particularly in terms of evaluation coefficients and error rates. To boost the model’s interpretability, this research developed and implemented a visualization technology scheme for the excellently performing models, integrating SHAP value analysis to vividly depict the feature space, prediction processes, and experimental outcomes, thereby aiding the exploration of Tg influencing factors. This research not only offers an innovative machine learning framework for Tg prediction in FKM but also enhances the model’s transparency and credibility using visualization technology, holding significant theoretical and practical implications for materials science research.

Accuracy versus precision in boosted top tagging with the ATLAS detector

Abstract The identification of top quark decays where the top quark has a large momentum transverse to the beam axis, known as top tagging, is a crucial component in many measurements of Standard Model processes and searches for beyond the Standard Model physics at the Large Hadron Collider. Machine learning techniques have improved the performance of top tagging algorithms, but the size of the systematic uncertainties for all proposed algorithms has not been systematically studied. This paper presents the performance of several machine learning based top tagging algorithms on a dataset constructed from simulated proton-proton collision events measured with the ATLAS detector at √ s = 13 TeV. The systematic uncertainties associated with these algorithms are estimated through an approximate procedure that is not meant to be used in a physics analysis, but is appropriate for the level of precision required for this study. The most performant algorithms are found to have the largest uncertainties, motivating the development of methods to reduce these uncertainties without compromising performance. To enable such efforts in the wider scientific community, the datasets used in this paper are made publicly available.

Scalable and Distributed Mathematical Modeling Algorithm Design and Performance Evaluation in Heterogeneous Computing Clusters

Scalable and Distributed Mathematical Modeling Algorithm Design and Performance Evaluation in Heterogeneous Computing Clusters

Nodule Detection and Generation on Chest X-Rays: NODE21 Challenge.

Pulmonary nodules may be an early manifestation of lung cancer, the leading cause of cancer-related deaths among both men and women. Numerous studies have established that deep learning methods can yield high-performance levels in the detection of lung nodules in chest X-rays. However, the lack of gold-standard public datasets slows down the progression of the research and prevents benchmarking of methods for this task. To address this, we organized a public research challenge, NODE21, aimed at the detection and generation of lung nodules in chest X-rays. While the detection track assesses state-of-the-art nodule detection systems, the generation track determines the utility of nodule generation algorithms to augment training data and hence improve the performance of the detection systems. This paper summarizes the results of the NODE21 challenge and performs extensive additional experiments to examine the impact of the synthetically generated nodule training images on the detection algorithm performance.

Comparative analysis of federated learning algorithms under non-IID data

With the increasing global attention to data privacy and security issues, how to effectively utilize distributed data while protecting personal privacy has become an important research topic. Federated learning (FL) aims to address data silos and privacy issues by enabling multiple devices or servers to train shared models in collaboration without submitting raw data to a central server. Although a variety of federated learning algorithms have been proposed, there is still a gap in the research on their performance differences under the same model. The goal of this study is to compare and analyze the performance differences of different federated learning algorithms on the same model through experiments. Based on the Fashion-MNIST dataset, this paper compares four commonly used federated learning algorithms in detail: Federal Averaging (FedAvg), Federated Stochastic Gradient Descent (FedSGD), Stochastic Controlled Averaging for Federated Learning (SCAFFOLD), and Federated Proximal (FedProx). The experimental results show that FedProx performs best in all evaluation indicators, followed by SCAFFOLD and FedAvg, while FedSGD performs the worst. These insights into algorithm performance with non-IID data inform practical application suitability and guide future research.

Isotopological remeshing and statistical shape analysis: Enhancing premolar tooth wear classification and simulation with machine learning

ObjectiveThe aim of this study was to evaluate the accuracy of a combined approach based on an isotopological remeshing and statistical shape analysis (SSA) to capture key anatomical features of altered and intact premolars. Additionally, the study compares the capabilities of four Machine Learning (ML) algorithms in identifying or simulating tooth alterations. Methods113 premolar surfaces from a multicenter database were analyzed. These surfaces were processed using an isotopological remeshing method, followed by a SSA. Mean Euclidean distances between the initial and remeshed STL files were calculated to assess deviation in anatomical landmark positioning. Seven anatomical features were extracted from each tooth, and their correlations with shape modes and morphological characteristics were explored. Four ML algorithms, validated through three-fold cross-validation, were assessed for their ability to classify tooth types and alterations. Additionally, twenty intact teeth were altered and then reconstructed to verify the method's accuracy. ResultsThe first five modes encapsulated 76.1% of the total shape variability, with a mean landmark positioning deviation of 10.4 µm (±6.4). Significant correlations were found between shape modes and specific morphological features. The optimal ML algorithms demonstrated high accuracy (>83%) and precision (>86%). Simulations on intact teeth showed discrepancies in anatomical features below 3%. ConclusionThe combination of an isotopological remeshing with SSA showed good reliability in capturing key anatomical features of the tooth. Clinical significanceThe encouraging performance of ML algorithms suggests a promising direction for supporting practitioners in diagnosing and planning treatments for patients with altered teeth, ultimately improving preventive care.

A stochastic ridesharing user equilibrium model with origin-destination-based ride-matching strategy

Ridesharing, as an emerging mode of modern urban transportation, has gained widespread popularity due to its ability to improve convenience and reduce the expenses in daily travel. To capture its impact on the network-wide traffic flow pattern, this paper develops a new variational inequality (VI) formulation for modeling the stochastic ridesharing user equilibrium (SRUE) problem considering an origin–destination (O-D) -based ride-matching strategy. The proposed SRUE problem is formulated based on a hierarchical logit choice model to depict the multi-stage decision-making (i.e., role choice, ride-matching, and route choice) of the ridesharing travelers. The SRUE model also reflects a two-sided market equilibrium considering the interaction effect between the ridesharing market and the road traffic equilibrium market. The logit choice model captures the ride-matching problem between the drivers and the passengers traveling on the same/different O-D pairs. Moreover, the lower- and upper-bound constraints for a ridesharing driver picking up passengers are established at the O-D level. For solving the SRUE model with O-D-based ride-matching constraints, the Lagrangian dual method (LDM) is adopted by incorporating the Barzilai-Borwein (BB) step size to update the dual variables, which accelerates the convergence of the LDM. Three numerical examples are conducted to illustrate the features of the proposed SRUE model and demonstrate the performance of the solution algorithm.

The Value of Radiomic Features of Primary Breast Tumor in STIR Sequences in Predicting Axilla Metastasis

Background: Detection of axillary metastases in breast cancer is critical for treatment options and prognosis. The aim of this study is to investigate the value of radiomic features obtained from short tau inversion recovery (STIR) sequences in magnetic resonance imaging (MRI) of primary tumor in breast cancer in predicting axillary lymph node metastasis (ALNM). Methods: Lesions of 165 patients with a mean age of 51.12 ±11 (range 28-82) with newly diagnosed invasive breast cancer who underwent breast MRI before treatment were manually segmented from STIR sequences in the 3D Slicer program in all sections. Machine learning (ML) analysis was performed using the extracted 851 features Python 3.11, Pycaret library program. Datasets were randomly divided into train (123, 80%) and independent test (63, 20%) datasets. The performances of ML algorithms were compared with area under curve (AUC), accuracy, recall, presicion and F1 scores. Results: Accuracy and AUC in the training set were in the range of 57 %-86 % and 0.50-0.95, respectively. The best model in the training set was the catBoost classifier with an AUC of 0.95 and 84% accuracy. The AUC, accuracy, recall, precision values and F1 score of the CatBoost classifier on the test set were 0.92, 84 %, 89 %, 85 %, 86 %, respectively. Conclusion: Radiomic features obtained from primary tumors on STIR sequences have the potential to predict ALNM in invasive breast cancer.

Sensitivity and specificity of International Classification of Diseases algorithms (ICD-9 and ICD-10) used to identify opioid-related overdose cases: A systematic review and an example of estimation using Bayesian latent class models in the absence of gold standards.

This study aimed to summarize validity estimates of International Classification of Diseases (ICD) codes in identifying opioid overdose (OOD) among patient data from emergency rooms, emergency medical services, inpatient, outpatient, administrative, medical claims, and mortality, and estimate the sensitivity and specificity of the algorithms in the absence of a perfect reference standard. We systematically reviewed studies published before December 8, 2023, and identified with Medline and Embase. Studies reporting sufficient details to recreate a 2 × 2 table comparing the ICD algorithms to a reference standard in diagnosing OOD-related events were included. We used Bayesian latent class models (BLCM) to estimate the posterior sensitivity and specificity distributions of five ICD-10 algorithms and of the imperfect coroner's report review (CRR) in detecting prescription opioid-related deaths (POD) using one included study. Of a total of 1990 studies reviewed, three were included. The reported sensitivity estimates of ICD algorithms for OOD were low (range from 25.0% to 56.8%) for ICD-9 in diagnosing non-fatal OOD-related events and moderate (72% to 89%) for ICD-10 in diagnosing POD. The last included study used ICD-9 for non-fatal and fatal and ICD-10 for fatal OOD-related events and showed high sensitivity (i.e. above 97%). The specificity estimates of ICD algorithms were good to excellent in the three included studies. The misclassification-adjusted ICD-10 algorithm sensitivity estimates for POD from BLCM were consistently higher than reported sensitivity estimates that assumed CRR was perfect. Evidence on the performance of ICD algorithms in detecting OOD events is scarce, and the absence of bias correction for imperfect tests leads to an underestimation of the sensitivity of ICD code estimates.

Federated learning: Impact of different algorithms and models on prediction results based on fashion-MNIST data set

The realm of federated learning is rapidly advancing amid the era of big data. Therefore, how to select a suitable federated learning algorithm to achieve realistic tasks has become particularly critical. In this study, we explore the impact of different algorithms and models on the prediction results of Federated Learning (FL) using the Fashion-MNIST data set. Federated Learning enhances data privacy and reduces latency by training models directly on local devices since it is a decentralized machine learning approach. We analyze the performance of several FL algorithms including Federated Averaging (FedAvg), Federated Stochastic Gradient Descent (FedSGD), Federated Proximal (FedProx), and SCAFFOLD. Our experiments reveal significant differences in accuracy and stability among these algorithms, highlighting their strengths and weaknesses in handling non-IID (Non-Independent and Identically Distributed) data. FedProx demonstrate superior performance in terms of accuracy and robustness, making them suitable for complex federated learning environments. These discoveries offer crucial insights for choosing suitable FL algorithms and models in practical applications.

High-resolution 3-D imaging of electron density perturbations using ultra-dense GNSS observation networks in Japan: an example of medium-scale traveling ionospheric disturbances

For the first time using computerized ionospheric tomography (CIT) and leveraging ultra-dense slant total electron content (STEC) measurements derived from two ground-based Global Navigation Satellite System (GNSS) receiver networks in Japan, we have reconstructed the 3-D field-aligned structure of nighttime medium-scale traveling ionospheric disturbances (MSTIDs) with high spatiotemporal resolution. The CIT algorithm focuses on electron density perturbation components, allowing for the imaging of disturbances with small amplitudes and scales. Slant TECs used for CIT are setup to consist of two components: the background derived from IRI-2016 model and TEC perturbations obtained by subtracting a 30-min running average from observations. The resolution is set to 0.25º in latitude and longitude, 10 km in altitude, 30 s in time. Simulations were conducted to assess the performance of the CIT algorithm, revealing that this technique has good fidelity by accurately reconstructing more than 80% of the electron density perturbations. The focus is on the nighttime event of July 4, 2022, when data were accessible. The reconstruction results show that the MSTIDs initially form at lower altitudes and subsequently develop to exhibit large amplitudes and scales that extend to higher altitudes, characterized by a well-defined frontal structure with electrodynamic signatures. These results are consistent with theories and snippets of observational evidence regarding electromagnetic-influenced MSTIDs, hence affirming the effectiveness of the developed CIT technique in probing of the variations in the 3-D structure of ionospheric electron density. This is expected to contribute to a compressive understanding of the underlying mechanisms of ionospheric inhomogeneities.Graphical

YOLO-Based Vehicle Detection: Literature Review

This research aims to evaluate the implementation of the You Only Look Once (YOLO) algorithm and its variants in the context of vehicle detection in traffic management systems. The importance of implementing intelligent transportation systems (ITS) in increasing transportation efficiency and reducing traffic problems such as congestion and accidents. The methodology used involves a critical review of current literature utilizing the YOLO algorithm for vehicle detection, with a focus on improving the accuracy of detection models. The research results show that the YOLO algorithm and its variants, such as YOLOv4 and YOLOv8, show a significant increase in vehicle detection accuracy reaching 90% in various environmental conditions. However, weaknesses in detecting small objects and in extreme lighting conditions still need further attention. This study also reviews several improvement approaches proposed in the literature, including the use of image augmentation techniques and the integration of deep learning models to improve the performance of the YOLO algorithm. The implementation of the YOLO algorithm in vehicle detection in intelligent transportation systems has great potential in increasing the efficiency and accuracy of traffic monitoring. This research provides recommendations for further development so that the YOLO algorithm can be better adapted to various environmental conditions and different types of data.

A hybrid genetic tabu search algorithm for distributed job-shop scheduling problems

The distributed job-shop scheduling problem (DJSP) is an extension of the traditional job-shop scheduling problem, which are composed of two sub-problems, assigning jobs to suitable factories and deciding the operation sequence on machines. To evaluate the performance of algorithms for solving DJSP, several famous benchmark instances have been proposed, and most of these instances have not been solved so far. This paper proposes a hybrid genetic tabu search algorithm (HGTSA) for solving DJSP. The proposed HGTSA combines the global search ability of the genetic algorithm (GA) and the local search ability of the tabu search (TS) well. In GA part, a crossover operation and a mutation operation are devised based on the critical factory. The two operations can effectively improve the discreteness of the population. In TS part, a tabu search procedure is performed on the critical factory. The procedure can effectively enhance the local search ability of HGTSA. For evaluating the performance of HGTSA, it has been compared with five classical algorithms on 240 benchmark instances. The computational results show the efficiency and effectiveness of HGTSA for solving DJSP. In particular, the proposed HGTSA updates the upper bounds for 235 out of these difficult instances.

The Role of Human Managers within Algorithmic Performance Management Systems: A Process Model of Employee Trust in Managers through Reflexivity

The Role of Human Managers within Algorithmic Performance Management Systems: A Process Model of Employee Trust in Managers through Reflexivity

Optimizing [formula omitted]-coverage in energy-saving wireless sensor networks based on the Elite Global Growth Optimizer

For energy-constrained wireless sensor networks, enhancing network fault tolerance and extending lifetime are of paramount importance. In this study, we focus on addressing the Minimum Power k-Coverage (MinPKC) problem. The objective of MinPKC is to minimize and balance the transmission (Tx) power by sensor nodes (SNs), thus extending network lifetime. Furthermore, MinPKC requires satisfying the constraints of k-coverage for the targets to enhance network fault tolerance. The problem is non-convex, multimodal, and NP-hard. To address this problem, we propose a novel metaheuristic algorithm named Elite Global Growth Optimizer (EGGO), which incorporates efficient global search and elite evolution strategies into the GO and utilizes time-varying multi-strategies for population updates. This study extensively compares EGGO with ten popular algorithms on 20 classic benchmark test functions using the Friedman and Wilcoxon rank-sum tests. The results indicate that EGGO significantly outperforms all the compared algorithms. Ablative experiments confirm the crucial role of the proposed mechanisms in EGGO for improving algorithm performance. Meanwhile, the study implements ten MinPKC methods based on comparative algorithms and compares them with the EGGO-based MinPKC method in terms of total Tx power, network lifetime, and simulation time. The experimental results indicate that the EGGO-based MinPKC method significantly increases the network lifetime by approximately 14.2%, 38.2%, and 1.9% compared to the second-best algorithm, and by 57%, 451%, and 84.8% compared to the GO, while balancing the Tx power among SNs and reducing the total Tx power by approximately 28%, 12.7%, and 0.5% compared to the second-best algorithm, and by 47.1%, 72%, and 30% compared to the GO, respectively, when k=1,2,3. Additionally, a detailed analysis is conducted on the impact of the number of SNs, the maximum target coverage capacity of each SN, as well as the effects of obstacles and their distribution on signal strength attenuation in reducing Tx power. The source code is available at https://github.com/iNet-WZU/EGGO.

A hybrid The Dwarf Mongoose Optimization Algorithm with Nelder- Mead method and its application for allocation reliability

The Dwarf Mongoose Optimization Algorithm (DMO) is inspired by the behaviour of Dwarf Mongoose which can strike the ideal balance throughout research between exploration and exploitation. In this article, we combine algorithms of the Dwarf Mongoose Optimization Algorithm and the Nelder-Mead Algorithm (DMONM). In addition, the statistically evaluated functions is utilized by calculating the average and the standard deviation values that are used to validate the suggested algorithm's performance. The experimental results are on high-efficiency optimization functions with various dimensions. The hybrid algorithm produces good, encouraging, and better outcomes than the original algorithms. The results show that the proposed algorithm could enhance the effects of DMO when it used to solve the optimization issues of the multi-objective reliability system

Evaluation in Neural Style Transfer: A Review

AbstractThe field of neural style transfer (NST) has witnessed remarkable progress in the past few years, with approaches being able to synthesize artistic and photorealistic images and videos of exceptional quality. To evaluate such results, a diverse landscape of evaluation methods and metrics is used, including authors' opinions based on side‐by‐side comparisons, human evaluation studies that quantify the subjective judgements of participants, and a multitude of quantitative computational metrics which objectively assess the different aspects of an algorithm's performance. However, there is no consensus regarding the most suitable and effective evaluation procedure that can guarantee the reliability of the results. In this review, we provide an in‐depth analysis of existing evaluation techniques, identify the inconsistencies and limitations of current evaluation methods, and give recommendations for standardized evaluation practices. We believe that the development of a robust evaluation framework will not only enable more meaningful and fairer comparisons among NST methods but will also enhance the comprehension and interpretation of research findings in the field.

Finding maximum independent set based on multi-stage simulated quantum adiabatic evolution

The ubiquitous nature of the maximum independent set (MIS) problem across diverse fields, including biology and chip design, presents a significant challenge. Recently, quantum algorithms have been recognized for their increased potential in addressing combinatorial optimization problems, including the MIS problem. Several quantum systems, including optical quantum circuits and Rydberg atomic systems, have been designed specifically for the MIS problem. However, elucidating the precise nature of instances that pose challenges for quantum algorithms constitutes a formidable task. In an effort to comprehend this issue, we endeavor to delineate the conditions under which instances become hard to solve by analyzing the Hamiltonian of the k-independent set (k-IS). Furthermore, we present two multi-stage strategies grounded in the simulated quantum adiabatic evolution (SQAE) of Kerr-nonlinear parametric oscillator (KPO) system for identifying the MIS in instances deemed hard to solve. More specifically, we present the algorithm leveraging SQAE of the KPO system for approximating the MIS. Our investigation of the SQAE algorithm reveals that optimal solutions are unattainable under two distinct conditions: (1) when the number of solutions is exponentially smaller to the problem’s scale, and (2) when the counts of non-optimal independent sets (IS) with m overlaps between the optimal k-IS (k>m) is exponentially smaller relative to the problem’s scale. With these two distinctive features, we propose two multi-stage SQAE strategies tailored for hard-to-solve instances. In the first approach, the SQAE algorithm is executed multiple times to identify a set of candidate vertices for the MIS. For each vertex or combinations of vertices in this set, we extract the subgraphs induced by the vertices not adjacent to them. Subsequently, the SQAE algorithm is applied to the subgraph to obtain the (k−1)-IS. In the second multi-stage strategy, we consider each vertex vi in the given graph, generating the subgraph induced by the vertices not adjacent to vi. The SQAE algorithm is then employed to find the (k−1)-IS of the subgraph. Experimental results demonstrate that both multi-stage strategies significantly enhance the performance of the SQAE algorithm for hard-to-solve instances.