778 results found.
Year: 2019, Volume: 19, Number: 1
... secondly k-means++ clustering [6] achieved a highest accuracy for English 88.76% and algorithm was used to validate the patterns obtained in the German 72.41% using maximum ... (SVM). These classifiers were trained on processing, machine learning, classification algorithms, the features and the resulted outcome were ensemble using supervised learning. various ...
Terms matched: 2 - Score: 88 - 1,249k
Year: 2009, Volume: 9, Issue: 1
... the sensitivity optimization problem by repeating the following two steps. to problem formulation, algorithm selection and usually 1) Candidate solutions are constructed using a converge to local ... values. At run-time, ACO representing the generation cost and/or the transmission algorithms try to update the pheromone values in such a way loss. The constraints ...
Terms matched: 2 - Score: 88 - 600k
Year: 2013, Volume: 13, Number: 1
... remanding to [17] for details. The idea is to use a recursive algorithm that performs the SVD-based approximations of multivariate functions suitable to simplify a MIMO problem ... NN into a collection of net architecture; 2) choice of the best learning algorithms, parallel-optimized SISO NNs. Experimental results will be training methodology and stopping criteria ...
Terms matched: 2 - Score: 87 - 689k
Year: 2015, Volume: 15, Number: 4
... in literature based on the voltamperometric method uses a microcontroller that implements the temperature control algorithm and also the power supply method [14]. Based on this method ... with the two control loops in analog domain and the macromodel used to test the algorithms. Figure 8. The digital implementation of a PI controller. I ref ...
Terms matched: 2 - Score: 87 - 1,502k
Year: 2022, Volume: 22, Number: 2
... random distribution of PVDGs in the distribution network can lead to An enhanced Sine Cosine Algorithm was proposed to obtain system security violations and congestion. Optimal allocation of the ... the mutation simulation in risking overvoltage. In this study, the proposed L other algorithms like the GA algorithm. However, the technique improves the voltage profile with ...
Terms matched: 2 - Score: 85 - 1,531k
Year: 2018, Volume: 18, Number: 4
... Electrical and Computer Engineering Volume 18, Number 4, 2018 An Efficient Deep Learning Algorithm for Fire and Smoke Detection with Limited Data Abdulaziz NAMOZOV, Young Im CHO ... color, making it a traditional data augmentation techniques and generative complicated task for those algorithms to achieve high adversarial networks. Experimental results show that the accuracy in detecting ...
Terms matched: 2 - Score: 84 - 1,980k
Year: 2017, Volume: 17, Number: 4
... of the art genetic multi-objective methodology and the obtained results are shown in Section 5 algorithm, response surface modelling methods and some and, finally, Section 6 concludes ... used for the optimization problems, which we used for the analysis and optimization of algorithms are presented in Section 4. Our research the COMSOL and MATLAB model of ...
Terms matched: 2 - Score: 84 - 1,377k
Year: 2017, Volume: 17, Number: 3
... an improved multi-frame image super-resolution (SR) class of regularization-based functions is the MAP algorithm is proposed. The objective function should be minimized consists of a data error ... . Total variation (TV) [15] is In SR literature, reconstruction-based algorithms are well another regularizer which overcomes the shortcoming of developed in spatial domain. ...
Terms matched: 2 - Score: 83 - 1,712k
Year: 2013, Volume: 13, Number: 1
... are given in [7]. of estimated modelling error is introduced into control algorithm to enhance the system steady-state accuracy. The In order to cope with ABS ... parameter given solution was verified in real experimental framework and uncertainties, robust non-linear control algorithms should be the obtained results were compared with the results of used. The ...
Terms matched: 2 - Score: 83 - 785k
Year: 2017, Volume: 17, Number: 3
... DPSO (RBF) [16]. Noted that the error penalty parameter C algorithm, comparative studies of fault diagnosis are controls the trade-off cost between the complexity ... thomht@ntu.edu.vn Abstract — This paper proposes a new differential particle supporting pattern recognition algorithms to get reliable swarm optimization (DPSO) method for obtaining optimum results. ...
Terms matched: 2 - Score: 83 - 1,291k