Sound classification using evolving ensemble models and Particle Swarm Optimization

被引:26
|
作者
Zhang, Li [1 ]
Lim, Chee Peng [2 ]
Yu, Yonghong [3 ]
Jiang, Ming [4 ]
机构
[1] Royal Holloway Univ London, Dept Comp Sci, Surrey TW20 0EX, England
[2] Deakin Univ, Inst Intelligent Syst Res & Innovat, Waurn Ponds, Vic 3216, Australia
[3] Nanjing Univ Posts & Telecommun, Coll Tongda, Nanjing, Peoples R China
[4] Univ Sunderland, Fac Technol, Sch Comp Sci, Sunderland, Durham, England
关键词
Sound classification; Evolutionary algorithm; Deep convolutional bidirectional long; short-term memory network and ensemble; classifier; CONVOLUTIONAL NEURAL-NETWORKS; GREY WOLF OPTIMIZER; FIREFLY ALGORITHM; RECOGNITION; REGRESSION;
D O I
10.1016/j.asoc.2021.108322
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Automatic sound classification attracts increasing research attention owing to its vast applications, such as robot navigation, environmental sensing, musical instrument classification, medical diagnosis, and surveillance. In this research, we propose an ensemble convolutional bidirectional Long Short Term Memory (CBiLSTM) network with optimal hyper-parameter selection for undertaking sound classification. We first transform each audio signal into a spectrogram representation using the Short time Fourier transform (STFT). A Particle Swarm Optimization (PSO) variant is subsequently proposed to optimize the learning rate, weight decay, numbers of filters and hidden units in the convolutional and BiLSTM layers, respectively, in order to extract effective spatial-temporal characteristics from the spectrogram inputs. To tackle the issue of stagnation in optimization, the proposed algorithm incorporates local exploitation using secant and Newton-Raphson methods, promising leader generation using regular and irregular super-ellipse formulae, and three-dimensional spherical search coefficients. Moreover, it takes into account multiple fused elite signals in conjunction with numerical analysis based exploitation to balance between diversification and intensification. A variety of CBiLSTM networks with distinctive optimized settings are devised. An ensemble model is then constructed by incorporating a set of three yielded networks based on a majority voting scheme. Evaluated using several audio data sets, our ensemble CBiLSTM networks outperform those with default and optimal settings identified by other search methods, existing deep architectures and state-of-the-art related studies. In addition to sound classification tasks, the proposed PSO algorithm also outperforms a number of classical and advanced search methods for solving diverse unimodal and multimodal benchmark functions with statistical significance. (C) 2021 Elsevier B.V. All rights reserved.
引用
收藏
页数:28
相关论文
共 50 条
  • [41] Novel Approach for Incremental Learning using Ensemble of SVMs with Particle Swarm Optimization
    Gupta, Aditya
    Gusain, Kunal
    Kumar, Deepika
    2016 11TH INTERNATIONAL CONFERENCE ON INDUSTRIAL AND INFORMATION SYSTEMS (ICIIS), 2016, : 426 - 430
  • [42] An Optimal Multi-view Ensemble Learning for High Dimensional Data Classification Using Constrained Particle Swarm Optimization
    Kumar, Vipin
    Minz, Sonajharia
    INFORMATION, COMMUNICATION AND COMPUTING TECHNOLOGY, 2017, 750 : 363 - 378
  • [43] Particle Swarm Optimization with Ensemble of Inertia Weight Strategies
    Shirazi, Muhammad Zeeshan
    Pamulapati, Trinadh
    Mallipeddi, Rammohan
    Veluvolu, Kalyana Chakravarthy
    ADVANCES IN SWARM INTELLIGENCE, ICSI 2017, PT I, 2017, 10385 : 140 - 147
  • [44] Evolving digital circuits using particle swarm
    Gudise, VG
    Venayagamoorthy, GK
    PROCEEDINGS OF THE INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS 2003, VOLS 1-4, 2003, : 468 - 472
  • [45] Plant Leaf Disease Detection and Classification Using Particle Swarm Optimization
    Yadav, Rishabh
    Rana, Yogesh Kumar
    Nagpal, Sushama
    MACHINE LEARNING FOR NETWORKING, 2019, 11407 : 294 - 306
  • [46] Training ANFIS Using Catfish-Particle Swarm Optimization for Classification
    Hassan, Norlida
    Ghazali, Rozaida
    Hussain, Kashif
    RECENT ADVANCES ON SOFT COMPUTING AND DATA MINING, 2017, 549 : 201 - 210
  • [47] Gender Classification Using Local Binary Pattern and Particle Swarm Optimization
    Khan, Sajid Ali
    Nazir, Muhammad
    Riaz, Naveed
    Hussain, M.
    Naveed, Nawazish
    EMERGING TRENDS AND APPLICATIONS IN INFORMATION COMMUNICATION TECHNOLOGIES, 2012, 281 : 73 - +
  • [48] Classification of Electroencephalogram Signals Using Wavelet Transform and Particle Swarm Optimization
    Ba-Karait, Nasser Omer
    Shamsuddin, Siti Mariyam
    Sudirman, Rubita
    ADVANCES IN SWARM INTELLIGENCE, ICSI 2014, PT II, 2014, 8795 : 352 - 362
  • [49] An evolutionary framework using particle swarm optimization for classification method PROAFTN
    Al-Obeidat, Feras
    Belacel, Nabil
    Carretero, Juan A.
    Mahanti, Prabhat
    APPLIED SOFT COMPUTING, 2011, 11 (08) : 4971 - 4980
  • [50] Classification of Leukemia Gene Expression Data Using Particle Swarm Optimization
    Liu, Yajie
    Shi, Xinling
    An, Zhenzhou
    2012 SIXTH INTERNATIONAL CONFERENCE ON GENETIC AND EVOLUTIONARY COMPUTING (ICGEC), 2012, : 241 - 244