• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.1
• /
• pp.133-151
• /
• 2019

In order to achieve faster content distribution speed and stronger fault tolerance, a P2P peer can connect to multiple peers in parallel and receive chunks of the data simultaneously. A critical issue in this environment is selecting a set of nodes participating in swarming sessions. Previous related researches only focus on performance metrics, such as downloading time or the round-trip time, but in this paper, we consider a new performance metric which is closely related to the network and propose a peer selection algorithm that produces the set of peers generating optimal worst link stress. We prove that the optimal algorithm is practicable and has the advantages with the experiments on PlanetLab. The algorithm optimizes the congestion level of the bottleneck link. It means the algorithm can maximize the affordable throughput. Second, the network load is well balanced. A balanced network improves the utilization of resources and leads to the fast content distribution. We also notice that if every client follows our algorithm in selecting peers, the probability is high that all sessions could benefit. We expect that the algorithm in this paper can be used complementary to existing methods to derive new and valuable insights in peer-to-peer networking.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.4
• /
• pp.1128-1145
• /
• 2022

Owing to the rapid development of information science, data analysis based on machine learning has become an interdisciplinary and strategic area. Marine predators algorithm (MPA) is a novel metaheuristic algorithm inspired by the foraging strategies of marine organisms. Considering the randomness of these strategies, an improved algorithm called co-evolutionary cultural mechanism-based marine predators algorithm (CECMPA) is proposed. Through this mechanism, search agents in different spaces can share knowledge and experience to improve the performance of the native algorithm. More specifically, CECMPA has a higher probability of avoiding local optimum and can search the global optimum quickly. In this paper, it is the first to use CECMPA to perform feature subset selection and optimize hyperparameters in support vector machine (SVM) simultaneously. For performance evaluation the proposed method, it is tested on twelve datasets from the university of California Irvine (UCI) repository. Moreover, the coronavirus disease 2019 (COVID-19) can be a real-world application and is spreading in many countries. CECMPA is also applied to a COVID-19 dataset. The experimental results and statistical analysis demonstrate that CECMPA is superior to other compared methods in the literature in terms of several evaluation metrics. The proposed method has strong competitive abilities and promising prospects.

• Korean Journal of Materials Research
• /
• v.33 no.4
• /
• pp.164-174
• /
• 2023

The bandgap characteristics of semiconductor materials are an important factor when utilizing semiconductor materials for various applications. In this study, based on data provided by AFLOW (Automatic-FLOW for Materials Discovery), the bandgap of a semiconductor material was predicted using only the material's compositional features. The compositional features were generated using the python module of 'Pymatgen' and 'Matminer'. Pearson's correlation coefficients (PCC) between the compositional features were calculated and those with a correlation coefficient value larger than 0.95 were removed in order to avoid overfitting. The bandgap prediction performance was compared using the metrics of R² score and root-mean-squared error. By predicting the bandgap with randomforest and xgboost as representatives of the ensemble algorithm, it was found that xgboost gave better results after cross-validation and hyper-parameter tuning. To investigate the effect of compositional feature selection on the bandgap prediction of the machine learning model, the prediction performance was studied according to the number of features based on feature importance methods. It was found that there were no significant changes in prediction performance beyond the appropriate feature. Furthermore, artificial neural networks were employed to compare the prediction performance by adjusting the number of features guided by the PCC values, resulting in the best R² score of 0.811. By comparing and analyzing the bandgap distribution and prediction performance according to the material group containing specific elements (F, N, Yb, Eu, Zn, B, Si, Ge, Fe Al), various information for material design was obtained.

• Journal of Communications and Networks
• /
• v.14 no.4
• /
• pp.434-442
• /
• 2012

Several gateway selection schemes have been proposed that select gateway nodes based on a single Quality of Service (QoS) path parameter, for instance path availability period, link capacity or end-to-end delay, etc. or on multiple non-QoS parameters, for instance the combination of gateway node speed, residual energy, and number of hops, for Mobile Ad hoc NETworks (MANETs). Each scheme just focuses on the ment of improve only a single network performance, i.e., network throughput, packet delivery ratio, end-to-end delay, or packet drop ratio. However, none of these schemes improves the overall network performance because they focus on a single QoS path parameter or on set of non-QoS parameters. To improve the overall network performance, it is necessary to select a gateway with stable path, a path with themaximum residual load capacity and the minimum latency. In this paper, we propose a gateway selection scheme that considers multiple QoS path parameters such as path availability period, available capacity and latency, to select a potential gateway node. We improve the path availability computation accuracy, we introduce a feedback system to updated path dynamics to the traffic source node and we propose an efficient method to propagate QoS parameters in our scheme. Computer simulations show that our gateway selection scheme improves throughput and packet delivery ratio with less per node energy consumption. It also improves the end-to-end delay compared to single QoS path parameter gateway selection schemes. In addition, we simulate the proposed scheme by considering weighting factors to gateway selection parameters and results show that the weighting factors improve the throughput and end-to-end delay compared to the conventional schemes.

• Journal of KIISE:Information Networking
• /
• v.29 no.1
• /
• pp.57-64
• /
• 2002

Since most of the current route selection algorithms use the shortest path algorithm, network resources can not be efficiently used also traffics be concentrated on specific paths resulting in congestgion. In this paper we propose the statistical route selections(SRS) algorithm which adopts a statistical mechanism to utilize the network resource efficiently and to avoid congestion. The SRS algorithm handles requests on demand and chooses a path that meets the requested bandwidth. With the advent of the MPLS it becomes possible to establish an explicit LSP which can be used for traffic load balancing. The SRS algorithm finds a set of link utilizations for route selection, computes link weights using statistical mechanism and finds the shortest path from the weights. Our statistical mechanism computes the mean and the variance of link utilizations and selects a route such that it can reduce the variance and the number of congested links and increase the utilization of network resources. Throughout the simulation, we show that the SRS algorithm performs better than other route selection algorithms on several metrics like the number of connection setup failures and the number of congested links.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.13 no.4
• /
• pp.1863-1871
• /
• 2012

Utilization of the former OSS was mainly focused on the usage of system applications such as operating system and DBMS. However, nowadays many companies are trying to make use of OSS based on the application rather than the system software. However, selection of base OSS is the most important to develop of application for utilizing OSS. The scope of existing OSS evaluation studies is covered the entire OSS quality. Thus existing studies of evaluation of OSS selection is insufficient. Also, the result of assessment is based on qualitative measurement rather than quantitative ones. In this paper, we derives only the indicators for selection among the existing OSS assessment indicators and suggests the assessment indicator that is capable of quantitative assessment in accordance with the characteristics of the project. The proposed assessment indicator is divided into an initial assessment indicator that can be assessed with only the information within the OSS community, and the detailed assessment indicator through metrics to make quantitative measurements possible. In this way, an objective basis can be provided through quantitative scores & indicators when selecting OSS.

• Journal of Internet Computing and Services
• /
• v.14 no.6
• /
• pp.25-32
• /
• 2013

Due to increasing competition of globalization and fast technological improvements the appropriate method for evaluating and selecting IS-personnel is one of the key factors for an organization's success. Personnel selection is a multi-criteria decision-making (MCDM) problem which consists of both qualitative and quantitative metrics. Although many articles have discussed various knowledge and skills IS personnel should possess, no specific model for IS personnel selection and evaluation, to our knowledge, has been published up to now. After reviewing the IS personnel's important characteristics, we propose an approach for categorizing the IS personnel based on their skills, ability, and knowledge during evaluation and selection process. Our proposed approach is derived from a model of neural network algorithm. We have adapted and implemented the fuzzy ART algorithm with Jaccard choice function. The result of an illustrative numerical example is proposed to demonstrate the easiness and effectiveness of our approach.

• Journal of Electrical Engineering and Technology
• /
• v.13 no.3
• /
• pp.1202-1211
• /
• 2018

In this paper, induction machine operation efficiency and torque is improved using Machine Learning based Gene Optimization (ML-GO) Technique is introduced. Optimized Genetic Algorithm (OGA) is used to select the optimal induction machine data. In OGA, selection, crossover and mutation process is carried out to find the optimal electrical machine data for induction machine design. Initially, many number of induction machine data are given as input for OGA. Then, fitness value is calculated for all induction machine data to find whether the criterion is satisfied or not through fitness function (i.e., objective function such as starting to full load torque ratio, rotor current, power factor and maximum flux density of stator and rotor teeth). When the criterion is not satisfied, annealed selection approach in OGA is used to move the selection criteria from exploration to exploitation to attain the optimal solution (i.e., efficient machine data). After the selection process, two point crossovers is carried out to select two crossover points within a chromosomes (i.e., design variables) and then swaps two parent's chromosomes for producing two new offspring. Finally, Adaptive Levy Mutation is used in OGA to select any value in random manner and gets mutated to obtain the optimal value. This process gets iterated till finding the optimal value for induction machine design. Experimental evaluation of ML-GO technique is carried out with performance metrics such as torque, rotor current, induction machine operation efficiency and rotor power factor compared to the state-of-the-art works.

• International Journal of Computer Science & Network Security
• /
• v.23 no.5
• /
• pp.148-162
• /
• 2023

Classification systems can significantly assist the medical sector by allowing for the precise and quick diagnosis of diseases. As a result, both doctors and patients will save time. A possible way for identifying risk variables is to use machine learning algorithms. Non-surgical technologies, such as machine learning, are trustworthy and effective in categorizing healthy and heart-disease patients, and they save time and effort. The goal of this study is to create a medical intelligent decision support system based on machine learning for the diagnosis of heart disease. We have used a mixed feature creation (MFC) technique to generate new features from the UCI Cleveland Cardiology dataset. We select the most suitable features by using Least Absolute Shrinkage and Selection Operator (LASSO), Recursive Feature Elimination with Random Forest feature selection (RFE-RF) and the best features of both LASSO RFE-RF (BLR) techniques. Cross-validated and grid-search methods are used to optimize the parameters of the estimator used in applying these algorithms. and classifier performance assessment metrics including classification accuracy, specificity, sensitivity, precision, and F1-Score, of each classification model, along with execution time and RMSE the results are presented independently for comparison. Our proposed work finds the best potential outcome across all available prediction models and improves the system's performance, allowing physicians to diagnose heart patients more accurately.

• The Korean Journal of Applied Statistics
• /
• v.35 no.4
• /
• pp.469-484
• /
• 2022

Since a large text corpus contains hundred-thousand unique words, text data is one of the typical large-dimensional data. Therefore, various feature selection methods have been proposed for dimension reduction. Feature selection methods can improve the prediction accuracy. In addition, with reduced data size, computational efficiency also can be achieved. The chi-square statistic and the information gain are two of the most popular measures for identifying interesting terms from text data. In this paper, we investigate the theoretical properties of the chi-square statistic and the information gain. We show that the two filtering metrics share theoretical properties such as non-negativity and convexity. However, they are different from each other in the sense that the information gain is prone to select more negative features than the chi-square statistic in imbalanced text data.