In humans, a brain abnormality is a serious illness. Cancer which is the greatest cause of mortality can develop from the tumour. Magnetic resonance imaging (MRI) is among a more extensively utilized medical imaging m...
详细信息
In humans, a brain abnormality is a serious illness. Cancer which is the greatest cause of mortality can develop from the tumour. Magnetic resonance imaging (MRI) is among a more extensively utilized medical imaging modalities in brain tumours, then it has become the primary diagnosing mechanism for the treatment and evaluation of brain tumours. Computer-assisted diagnosis has become a requirement due to the exponential expansion in the quantity of MRIs acquired because of these programmes. Computer-assisted diagnosis strategies created to increase detection without many systematized readings failed to produce significant improvements in performance measurements. In this regard, the usage of deep learning-based automatic image processing algorithms appears to be a viable route for identifying brain cancer. In this research, introduce a catswarmoptimization (CSO) algorithm based upon a convolutional neural network (CNN) model utilized to segmentation in a classification of brain tumour. Results of experiments on MRI images using the BRATS dataset show that the CSO algorithm-CNN model achieved high-performance in term of 98% of accuracy, precision, specificity, sensitivity and F-score in the proposed classification task when compared to other classification approaches like support vector machines (SVM) as well as back propagation neural networks (BPNN).
Well placement optimization is a complex and time-consuming task. An efficient and robust algorithm can improve the optimization efficiency. In this work, we propose a meta-optimized hybrid catswarm mesh adaptive dir...
详细信息
Well placement optimization is a complex and time-consuming task. An efficient and robust algorithm can improve the optimization efficiency. In this work, we propose a meta-optimized hybrid catswarm mesh adaptive direct search (O-CSMADS) algorithm for well placement optimization. By coupling catswarmoptimization (CSO) algorithm, Mesh Adaptive Direct Search (MADS) algorithm, and Particle swarmoptimization (PSO) meta-optimization approach, O-CSMADS has global search ability and local search ability. We perform detailed comparisons of optimization performances between O-CSMADS, hybrid catswarm mesh adaptive direct search (CSMADS) algorithm, CSO, and MADS in three different examples. Results show that O-CSMADS algorithm outperforms stand-alone CSO, MADS, and CSMADS. Besides, optimal controlling parameters are not same for different problems, which indicates that the optimization of algorithmic parameters is necessary. The proposed method also shows great potential for other petroleum engineering optimization problems, such as well type optimization and joint optimization of well placement and control. (C) 2018 Elsevier Ltd. All rights reserved.
In view of solving multi-objective path planning in the static environment,there are some faults for ant colony optimization(ACO),such as the long computation and easy to fall into local *** solve these problems,the A...
详细信息
In view of solving multi-objective path planning in the static environment,there are some faults for ant colony optimization(ACO),such as the long computation and easy to fall into local *** solve these problems,the ACO based on catswarmoptimization(CSO) algorithm searching model(CSOACO) is *** this algorithm,the introduction of CSO algorithm search pattern realizes the local search in the current solution for ant colony individuals,which not only enrich the diversity of solution,but improve the accuracy of the ***,the new algorithm is simulated in MATLAB for picking robot multi-objective path planning *** the simulation analysis,not only set the parameters,but compare CSOACO with other *** results show that the algorithm can accelerate the convergence speed,search to the global optimal solution and realize the multiobjective path planning of picking robot.
Aiming at the time-varying incident angle of the coherent signal source, this paper studies the effect of updating the data covariance matrix and the value of the forgetting factor on the tracking result. On the basis...
详细信息
ISBN:
(纸本)9781728144801
Aiming at the time-varying incident angle of the coherent signal source, this paper studies the effect of updating the data covariance matrix and the value of the forgetting factor on the tracking result. On the basis of this, the idea of group intelligent search using the catswarmoptimization (CSO) algorithm is used to calculate the maximum likelihood (ML) algorithm. The huge amount of problems is improved, and a catswarmoptimization-maximum likelihood (CSO-ML) algorithm dynamic DOA tracking method is proposed. This method does not need to pre-process the sample covariance matrix using decoherence technology, and can directly process the coherent signal source, and it is still effective in the case that the algorithm such as projection approximation subspace tracking (Past) algorithm with a small forgetting factor value fails. Simulation experiments show that this method has better tracking accuracy under the condition of low signal-to-noise ratio (SNR) and small snapshots and can directly deal with coherent signal sources.
In wireless communication systems, maximizing spectral efficiency and enhancing link reliability are key objectives. One effective solution is the combination of Orthogonal Frequency Division Multiplexing (OFDM) and M...
详细信息
In wireless communication systems, maximizing spectral efficiency and enhancing link reliability are key objectives. One effective solution is the combination of Orthogonal Frequency Division Multiplexing (OFDM) and Multiple-Input Multiple-Output (MIMO) techniques. OFDM divides the frequency band into non-overlapping sub-bands, enabling parallel data transmission. This helps overcome the limitations of traditional wireless communication systems with high-rate input data streams. MIMO-OFDM has emerged as a promising technology for achieving high data rates and robustness in wireless communications by supporting multiple inputs and outputs. However, optimizing both spectral efficiency and performance in rapidly fading channels can be challenging. To address these issues, we propose Time Frequency Training Orthogonal Frequency Division Multiplexing (TFT-OFDM). This approach utilizes group pilots for spectrum efficiency and shared channel estimates to maintain system performance stability. Channel estimation plays a critical role in TFT-OFDM, and both Least Square (LS) and Minimum Mean Square Error (MMSE) methods are considered. LS estimation has a simple approach but lower performance, while MMSE significantly reduces mean square error at the cost of computational complexity. To estimate the channel in TFT-OFDM, joint time and frequency channel estimation techniques are employed. This involves using sophisticated algorithms to optimize parameters such as maximum or minimum function values in the solution space. The intelligent algorithms, specifically the Bald Eagle Search optimization (BESOA) technique and the cat swarm optimization algorithm (CSOA), are utilized to estimate the channel efficiently in spectrally efficient MIMO-OFDM wireless networks. Performance improvement is observed with the intelligent algorithms. For instance, at a signal-to-noise ratio (SNR) of 15 dB, the proposed BESOA method achieves a Bit Error Rate (BER) of 10-6, while TFT-OFDM with group pi
暂无评论