Article
Environmental Sciences
Long Cui, Jiahua Zhang, Zhenjiang Wu, Lan Xun, Xiaopeng Wang, Shichao Zhang, Yun Bai, Sha Zhang, Shanshan Yang, Qi Liu
Summary: Wetlands in the Yellow River Delta are important and vulnerable due to tidal action and sediment deposits. A object-oriented approach with feature preference machine learning was used to classify the wetlands. A superpixel segmentation method using the watershed algorithm improved the classification accuracy. The random forest classifier combining superpixel segmentation and feature selection methods outperformed other pixel-based machine learning methods with a 91.74% overall accuracy and a kappa coefficient of 0.9078.
ENVIRONMENTAL SCIENCE AND POLLUTION RESEARCH
(2023)
Article
Computer Science, Information Systems
Xiangyuan Gu, Jianguo Chen, Guoqiang Wu, Kun Wang, Jiaxing Wang
Summary: Due to the limitation of using only one metric or comparing two metrics separately to measure redundant features, some existing feature subset selection algorithms fail to achieve the desired performance. To address this issue, a feature subset selection algorithm called symmetric uncertainty and interaction factor (SUIF) is proposed. SUIF evaluates relevant features using symmetric uncertainty and removes irrelevant features. It then uses graph theory to process relevant features, removes edges with lower weights, and clusters features using the Louvain community detection algorithm. Finally, SUIF evaluates features in each cluster using symmetric uncertainty, interaction factor, and equal interval division and ranking, and eliminates redundant features. Experimental results show that SUIF outperforms other algorithms in terms of feature selection performance.
MULTIMEDIA TOOLS AND APPLICATIONS
(2023)
Review
Computer Science, Artificial Intelligence
Carlos Villa-Blanco, Concha Bielza, Pedro Larranaga
Summary: Real-world problems often have high feature dimensionality, making it difficult to model and analyze the data. Feature subset selection (FSS) techniques can be used to reduce irrelevant or redundant information, improving the speed and performance of building models. This review focuses on incremental FSS algorithms that can efficiently handle large volumes of data received sequentially. Different strategies, such as updating feature weights incrementally, applying information theory, or using rough set-based FSS, are discussed, along with various supervised and unsupervised learning tasks where FSS is applicable.
ARTIFICIAL INTELLIGENCE REVIEW
(2023)
Article
Geochemistry & Geophysics
Zhaocong Wu, Zhao Yan
Summary: This letter proposes a novel end-to-end HSI local feature descriptor network called HyperDesc, which implements a true band selection module by turning band selection into a differentiable sampling operation. Experiments demonstrate that the spectral information provided by selected bands can boost the performance of the descriptor.
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS
(2022)
Article
Mathematics, Interdisciplinary Applications
Tayebeh Iloon, Ramin Barati, Hamid Azad
Summary: This paper proposes a supervised model for epileptic seizure detection using a Siamese network and a support vector machine classifier. Through feature extraction and transformation, the model achieves a classification performance with 100% accuracy, which is of positive significance for doctors in detecting epileptic seizure activity.
Article
Computer Science, Hardware & Architecture
Shanshan Xie, Yan Zhang, Danjv Lv, Xu Chen, Jing Lu, Jiang Liu
Summary: Feature selection plays a crucial role in pattern recognition and data mining. This paper proposes an improved maximal relevance and minimal redundancy (ImRMR) feature selection method based on feature subset, which adjusts the weights of measurement criteria and generates candidate feature subsets using equal grouping and incremental search methods. Experimental results demonstrate that ImRMR can effectively remove irrelevant and redundant features, leading to improved classification performance.
JOURNAL OF SUPERCOMPUTING
(2023)
Article
Computer Science, Artificial Intelligence
Nicolas Garcia-Pedrajas, Gonzalo Cerruela-Garcia
Summary: Feature subset selection is a common procedure in machine learning, with methods classified as embedded, filter, and wrapper. Wrappers achieve better classification performance but suffer from scalability issues. Filters are typically faster and more applicable for large datasets. This paper proposes a new method called MABUSE that optimizes margins from a filter perspective for feature selection. Experimental validation shows that it outperforms other algorithms in classification and reduction tasks, with similar computational cost.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Biochemistry & Molecular Biology
Leqi Tian, Wenbin Wu, Tianwei Yu
Summary: Random Forest (RF) is a popular machine learning method for classification and regression tasks, and it performs well under low sample size situations. However, there are issues with gene selection using RF as the important genes are usually scattered on the gene network, which conflicts with the biological assumption of functional consistency. To address this issue, we propose the Graph Random Forest (GRF) method, which incorporates external topological information to identify highly connected important features. The algorithm achieves equivalent classification accuracy to RF while selecting interpretable feature sub-graphs.
Article
Computer Science, Artificial Intelligence
Zhehuang Huang, Jinjin Li
Summary: This study proposes a new uncertain measure called multi-level granularity entropy, which explores the granularity structure and entropy of fuzzy covering through different fuzzy neighborhoods. Based on this, several variants of the multi-level granularity entropy are proposed to reflect the change of uncertain information. Finally, a dimensionality reduction method is implemented using the multi-level entropy, and a forward feature selection algorithm is developed. Extensive numerical experiments show that the presented model performs better in feature learning and outperforms some representative feature selection algorithms in terms of classification accuracy and the number of selected features.
ARTIFICIAL INTELLIGENCE REVIEW
(2023)
Article
Computer Science, Information Systems
Cun Ji, Mingsen Du, Yanxuan Wei, Yupeng Hu, Shijun Liu, Li Pan, Xiangwei Zheng
Summary: Time series classification is widely used in various domains, including EEG/ECG classification, device anomaly detection, and speaker authentication. Despite the existence of many methods, selecting intuitive temporal features for accurate classification remains a challenge. Therefore, this paper proposes a new method called TSC-RTF, which utilizes random temporal features, and shows that it can compete with state-of-the-art methods.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2023)
Article
Computer Science, Information Systems
Zhijun Chen, Qiushi Chen, Yishi Zhang, Lei Zhou, Junfeng Jiang, Chaozhong Wu, Zhen Huang
Summary: In this paper, a novel feature subset selection algorithm called CFSRCA is proposed, which effectively selects candidate class-relevant features and representative features, and the experimental results validate its effectiveness.
COMPUTER COMMUNICATIONS
(2021)
Article
Computer Science, Information Systems
Lizeng Gong, Shanshan Xie, Yan Zhang, Mengyao Wang, Xiaoyan Wang
Summary: This paper proposes a feature selection method based on factor analysis, which improves the classification accuracy and reduces the dimensionality by removing redundancy and obtaining the optimal feature subsets.
Article
Computer Science, Artificial Intelligence
Xiangyuan Gu, Jichang Guo
Summary: The paper proposes a new feature subset selection algorithm based on equal interval division and three-way interaction information to address the issues existing in some current algorithms. Experimental results demonstrate that the proposed algorithm can achieve better feature selection performance.
Article
Computer Science, Information Systems
Joyce A. Ayoola, Tokunbo Ogunfunmi
Summary: Advancements in computer-aided tools have improved accurate breast cancer prediction models, reducing the mortality rate. Random forest predictor and genetic algorithm are key methods in achieving high accuracy and effective feature selection. This paper proposes hybridized genetic algorithm models for breast cancer prediction, considering the order of feature selection algorithms, and compares their performance with other learning models. The hybridized Genetic Algorithm with Fisher_Score (GA + Fisher_Score) model shows promising results with 99.12% accuracy, outperforming other hybridized genetic algorithm models.
Article
Multidisciplinary Sciences
Abeer Elkhouly, Allan Melvin Andrew, Hasliza A. Rahim, Nidhal Abdulaziz, Mohd Fareq Abd Malek, Shafiquzzaman Siddique
Summary: In this study, a machine learning solution based on unsupervised spectral clustering is introduced to classify audiograms according to their shapes. The proposed ML algorithm outperforms existing models with higher accuracy, precision, recall, specificity, and F-score values. This work presents a novel ML technique that can potentially change the existing practices in classifying audiograms.
SCIENTIFIC REPORTS
(2023)
Article
Engineering, Electrical & Electronic
Dayakshini Sathish, Surekha Kamath, Keerthana Prasad, Rajagopal Kadavigere, Roshan J. Martis
SIGNAL IMAGE AND VIDEO PROCESSING
(2017)
Article
Engineering, Aerospace
R. Devaraj, K. Sankarasubramanian, Surekha Kamath
ADVANCES IN SPACE RESEARCH
(2013)
Article
Engineering, Chemical
Surekha Kamath, V. I. George, Sudha Vidyasagar
CANADIAN JOURNAL OF CHEMICAL ENGINEERING
(2009)
Article
Engineering, Electrical & Electronic
Surekha Kamath, V. I. George, Sudha Vidyasagar
IETE JOURNAL OF RESEARCH
(2009)
Article
Engineering, Electrical & Electronic
Surekha Kamath, V. I. George, Sudha Vidyasagar
IETE JOURNAL OF RESEARCH
(2009)
Article
Biotechnology & Applied Microbiology
Kavitha Govarthanan, Piyush Kumar Gupta, Bamadeb Patra, Deepa Ramasamy, Binita E. Zipporah, Vineeta Sharma, Rajesh Yadav, Pavitra Kumar, Dayakshini Sathish, Rama Shanker Verma
Summary: This study predicted the propensity of mesenchymal stem cells (MSCs) to differentiate towards the cardiovascular lineage using global methylome profiling and confirmed their higher propensity in vitro. It also identified activated pathways related to the cardiovascular lineage and discovered upregulated expression of certain cardiac-specific transcription factors in MSCs.
Article
Engineering, Electrical & Electronic
Vaidehi Nayantara Pattwakkar, Surekha Kamath, Manjunath Kanabagatte Nanjundappa, Rajagopal Kadavigere
Summary: This article presents a deep learning-based method for liver segmentation, which accurately segments the liver and liver tumors. By using the SegNet model and the K-means clustering method combined with image enhancement techniques, good segmentation results were obtained, and the method demonstrated stability across all phases.
INTERNATIONAL JOURNAL OF IMAGING SYSTEMS AND TECHNOLOGY
(2023)
Proceedings Paper
Materials Science, Multidisciplinary
T. Raja, R. Prabhakaran, D. Praveen Kumar, D. Sathish
Summary: This study successfully improves the mechanical and tribological properties of aluminum-based composites by adding reinforcement materials such as B4C and MoS2. The addition of reinforcing materials increases the hardness and tensile strength of the composites, while reducing the coefficient of friction. Microscopic investigation reveals that the reinforcement particles are uniformly distributed in the matrix alloy of aluminum solid solutions.
MATERIALS TODAY-PROCEEDINGS
(2022)
Proceedings Paper
Engineering, Electrical & Electronic
Vidya S. Rao, V. George, Surekha Kamath, C. Shreesha
FIFTH INTERNATIONAL CONFERENCE ON WIRELESS AND OPTICAL COMMUNICATIONS
(2017)
Proceedings Paper
Automation & Control Systems
Vidya S. Rao, V. I. George, Surekha Kamath, C. Shreesha
2015 10TH ASIAN CONTROL CONFERENCE (ASCC)
(2015)
Proceedings Paper
Computer Science, Information Systems
M. V. Dileep, Surekha Kamath, Vishnu G. Nair
ELEVENTH INTERNATIONAL CONFERENCE ON COMMUNICATION NETWORKS, ICCN 2015/INDIA ELEVENTH INTERNATIONAL CONFERENCE ON DATA MINING AND WAREHOUSING, ICDMW 2015/NDIA ELEVENTH INTERNATIONAL CONFERENCE ON IMAGE AND SIGNAL PROCESSING, ICISP 2015
(2015)