Article
Biochemistry & Molecular Biology
Rimsha Asad, Saif Ur Rehman, Azhar Imran, Jianqiang Li, Abdullah Almuhaimeed, Abdulkareem Alzahrani
Summary: Brain tumors can cause damage to the brain and surrounding tissues, blood vessels, and nerves if not treated promptly. Early detection is crucial to prevent fatal outcomes. Manual detection is challenging due to variations in tumor characteristics, thus an automatic system using a deep convolutional neural network is proposed. The system achieved high accuracy on the brain-tumor dataset and outperformed baseline methods, demonstrating its potential for application in other diseases.
Article
Chemistry, Analytical
Chan-Il Kim, Seok-Min Hwang, Eun-Bin Park, Chang-Hee Won, Jong-Ha Lee
Summary: The study proposed a computer-aided diagnostic algorithm for the classification of malignant melanoma and benign skin tumors using deep learning techniques. By employing U-Net model and convolutional neural networks, the algorithm achieved high accuracy in skin lesion classification.
Article
Computer Science, Artificial Intelligence
Changda Xing, Yuhua Cong, Chaowei Duan, Zhisheng Wang, Meiling Wang
Summary: This article presents a novel deep network, DIKS, for the classification of hyperspectral images. By using irregular convolutional kernels and self-expressive property, the network can adaptively compute feature maps to describe the characteristics of different object classes. The introduced self-expression theory helps produce more discriminative features through clustering. Experimental results show that this method outperforms state-of-the-art algorithms in classification performance.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Mathematics
Qi Zhang, Zuobin Ying, Jianhang Zhou, Jingzhang Sun, Bob Zhang
Summary: This paper proposes a broad learning model with a dual feature extraction strategy (BLM_DFE) that utilizes kernel principal component analysis (KPCA) to preprocess input data and extract effective low-dimensional features. The model simplifies the architecture of broad learning and achieves superior recognition performance.
Article
Computer Science, Artificial Intelligence
Pei Li, Wenlin Zhang, Chengjun Lu, Rui Zhang, Xuelong Li
Summary: A novel robust kernel principal component analysis method with optimal mean (RKPCA-OM) is proposed to enhance the robustness of KPCA by automatically eliminating the optimal mean. The theoretical proof guarantees the convergence of the algorithm and the obtained optimal subspaces and means. Exhaustive experimental results validate the superiority of the proposed method.
Article
Engineering, Biomedical
Dongfang Gao, Xu Zhang, Chao Zhou, Wei Fan, Tianyi Zeng, Qian Yang, Jianmin Yuan, Qiang He, Dong Liang, Xin Liu, Yongfeng Yang, Hairong Zheng, Zhanli Hu
Summary: This study proposes a new kernel reconstruction method that effectively reduces noise in reconstructed images and improves the image quality and accuracy of tumor regions by extracting autocorrelation texture features.
PHYSICS IN MEDICINE AND BIOLOGY
(2021)
Article
Engineering, Multidisciplinary
Junxian Shen, Feiyun Xu
Summary: The health monitoring system for equipment is crucial for industrial production. This study proposes a feature selection and fusion method based on poll mode and optimized Weighted Kernel Principal Component Analysis (WKPCA) method to dig deeper for effective features and improve the separability of fault samples.
Article
Geochemistry & Geophysics
Jianwei Zheng, Yuchao Feng, Cong Bai, Jinglin Zhang
Summary: This study proposed a mixed CNN with covariance pooling approach for HSI classification. By mixing 3-D and 2-D convolutions, utilizing covariance pooling technique, and employing PCA-involved strategies, the proposed model effectively improved classification accuracy and stability, especially in scenarios with limited sample size.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2021)
Article
Thermodynamics
Aosong Liang, Yunpeng Hu, Guannan Li
Summary: This study investigates the impact of different anomaly detection methods on chiller sensor fault detection and validates the results using an experimental dataset. The results show that anomaly detection methods can improve the quality of the original training data and KPCA has higher fault detection efficiency compared to PCA. IF-KPCA and Kmeans-KPCA further enhance the fault detection efficiency on top of KPCA.
INTERNATIONAL JOURNAL OF REFRIGERATION
(2023)
Article
Computer Science, Information Systems
De-He Yang, Xin Zhou, Xiu-Ying Wang, Jian-Ping Huang
Summary: The discrimination of source depth in micro-earthquake monitoring is crucial, and this study explores the use of machine learning techniques, showing that deep learning outperforms traditional classification methods in distinguishing source depth.
INFORMATION SCIENCES
(2021)
Article
Materials Science, Composites
Hua Wang, Chen Yan, Junyang Yu, Kristina Warmefjord
Summary: The paper presents a clamping force integrated computer aided tolerancing method for composite assembly, which modifies the probability distribution of clamping forces and coordinates them based on the main deformation mode to satisfy the coaxial tolerance requirements of composite components.
JOURNAL OF COMPOSITE MATERIALS
(2021)
Article
Forestry
Xiaofang Sun, Liping Sun, Yinglai Huang
Summary: The study proposed an improved convolutional neural network (CNN) that can automatically learn feature information of smoke images in the early stages of a forest fire with high recognition rate, without the need for complex manual feature extraction. By performing kernel principal component analysis and applying optimization strategies, the efficiency of smoke discrimination was accelerated, and the robustness of the CNN was enhanced while reducing overfitting.
JOURNAL OF FORESTRY RESEARCH
(2021)
Article
Computer Science, Theory & Methods
Jinfu Chen, Yuhao Chen, Saihua Cai, Shang Yin, Lingling Zhao, Zikang Zhang
Summary: Abnormal network traffic detection is an important technology for ensuring cybersecurity by identifying malicious attacks based on the behavior of network traffic. However, existing feature extraction methods for abnormal network traffic detection lack efficiency and are unsuitable for diverse and complex network traffic. To address this, this paper proposes an optimized feature extraction algorithm called LD-KPCA, which combines Linear Discriminant Analysis (LDA) and Kernel Principal Component Analysis (KPCA) to improve the efficiency and effectiveness of feature extraction in abnormal network traffic detection. Experimental results demonstrate that the LD-KPCA algorithm achieves high precision, recall, and F1-measure in detecting abnormal network traffic.
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
(2023)
Article
Computer Science, Interdisciplinary Applications
Kaidi Liu, Zijian Zhao, Pan Shi, Feng Li, He Song
Summary: Surgical tool detection is important for computer-assisted surgery, but data shortage and the balance between accuracy and speed remain challenges. This study manually annotated a new dataset and proposed an enhanced feature-fusion network (EFFNet) for real-time surgical tool detection. The method achieved high accuracy and met the real-time standard.
JOURNAL OF COMPUTATIONAL DESIGN AND ENGINEERING
(2022)
Article
Engineering, Electrical & Electronic
Congxin Han, Zhiyong Wang, Aixia Tang, Hongxin Gao, Fengyi Guo
Summary: A method based on KPCA and FA-SVM was proposed to recognize arc fault features under complicated harmonic conditions, effectively improving identification accuracy and stability.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2021)