Sound classification using evolving ensemble models and Particle Swarm Optimization

Zhang, Li; Lim, Chee Peng; Yu, Yonghong; Jiang, Ming

doi:10.1016/j.asoc.2021.108322

Cited by 29 publications

(10 citation statements)

References 83 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For the ICBHI dataset, as indicated in Table 25 , most existing studies, e.g., [ 13 , 22 , 23 , 24 , 26 ], categorized the six disease classes into healthy/unhealthy (two-class) or healthy/chronic/non-chronic (three-class) cases and performed binary or three-class predictions. In comparison with such classification tasks, our ensemble model performed a comparatively more challenging task for the identification of six respiratory abnormalities and achieved competitive performances.…”

Section: Discussionmentioning

confidence: 99%

“…Moreover, most of the existing studies, such as [ 13 , 20 , 22 , 23 ], employed a random train–test split, instead of a subject-independent split, which may have audio clips from the same subjects allocated in both training and test sets, although the recordings were collected from different chest locations. In contrast, Zhang et al [ 26 ] utilized a subject-independent split as those used in this research but their work conducted a three-class classification to identify healthy/chronic/non-chronic cases. García-Ordás et al [ 24 ] also classified healthy/chronic/non-chronic cases using a 10-fold cross-validation with a random split.…”

Section: Discussionmentioning

confidence: 99%

“…In addition, most studies for the ICBHI dataset, such as Perna and Tagarelli [ 13 ], Wall et al [ 20 ], and Wall et al [ 22 ], adopted RNN models, e.g., LSTM, BiLSTM, or BiGRU, for lung abnormality detection, while a CNN with five convolutional layers and a CRNN model were used by García-Ordás et al [ 24 ] and Zhang et al [ 26 ], respectively. In comparison with these works, our ensemble model incorporates all the above different types of networks (i.e., BiLSTM, BiGRU, CNN, and CRNN), each with PSO-optimized learning hyper-parameters.…”

Section: Discussionmentioning

confidence: 99%

“…For example, Boddapati et al [ 25 ] conducted environmental sound identification using AlexNet and GoogLeNet. Zhang et al [ 26 ] exploited an ensemble of CRNN models with hyper-parameter selection for respiratory, heart, and environmental sound classification, while 1D CNNs were utilized by Li et al [ 27 ] and Xiao et al [ 28 ] for heart abnormality classification using audio inputs. We summarize the aforementioned studies in Table 1 .…”

Section: Related Workmentioning

confidence: 99%

See 3 more Smart Citations

A Deep Ensemble Neural Network with Attention Mechanisms for Lung Abnormality Classification Using Audio Inputs

Wall

Zhang

et al. 2022

Sensors

Self Cite

View full text Add to dashboard Cite

Medical audio classification for lung abnormality diagnosis is a challenging problem owing to comparatively unstructured audio signals present in the respiratory sound clips. To tackle such challenges, we propose an ensemble model by incorporating diverse deep neural networks with attention mechanisms for undertaking lung abnormality and COVID-19 diagnosis using respiratory, speech, and coughing audio inputs. Specifically, four base deep networks are proposed, which include attention-based Convolutional Recurrent Neural Network (A-CRNN), attention-based bidirectional Long Short-Term Memory (A-BiLSTM), attention-based bidirectional Gated Recurrent Unit (A-BiGRU), as well as Convolutional Neural Network (CNN). A Particle Swarm Optimization (PSO) algorithm is used to optimize the training parameters of each network. An ensemble mechanism is used to integrate the outputs of these base networks by averaging the probability predictions of each class. Evaluated using respiratory ICBHI, Coswara breathing, speech, and cough datasets, as well as a combination of ICBHI and Coswara breathing databases, our ensemble model and base networks achieve ICBHI scores ranging from 0.920 to 0.9766. Most importantly, the empirical results indicate that a positive COVID-19 diagnosis can be distinguished to a high degree from other more common respiratory diseases using audio recordings, based on the combined ICBHI and Coswara breathing datasets.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

A Deep Ensemble Neural Network with Attention Mechanisms for Lung Abnormality Classification Using Audio Inputs

Wall

Zhang

et al. 2022

Sensors

Self Cite

View full text Add to dashboard Cite

show abstract

“…However, it is a problem that is also significant to resource use [57]. Furthermore, this data may contain a variety of noisy material and less useful features that are not essential for detecting an attack pattern [58]. As a result, the following procedure is utilised to maximize data quality and dimension reduction precision must be attained and effectiveness in terms of memory and time use.…”

Section: Fig 1 Conceptual System Block Flowmentioning

confidence: 99%

Mechanism for Extracting Features Using Particle Swarm Optimization for Intrusion Detection Systems

Sivagaminathan

Sharma

Henge

2023

Preprint

View full text Add to dashboard Cite

The task of ensuring cyber-security has grown increasingly challenging given the concerning expansion of Computing connection and furthermore, there are a large number of computer-related applications available. It also needs a strong defense mechanism towards various cyber-attacks. Identifying irregularities and dangers in such a computer Security measures (IDS) have been established to aid with information security. Particularly, ML approaches are a subset of artificial intelligence (ai). (AI), a useful data-driven anti - malware system was developed. Two alternative intrusion detection (ID) classification reaches were compared in this study, each with its own set of use cases. Before using the two classifiers for classification, the Particle Swarm Optimization (PSO) approach ware used for reduce dimensionality. The classification meets used to characterize network anomalies were studied in this study. PSO + ANN (Artificial neural network), PSO plus Decision Tree and PSO plus K-Nearest Neighbor are the three classifiers used. The Knowledge discovery in databases 99 datasets was used to corroborate the identification techniques' findings. On the result of the implementation, successful metrics like as the following metrics were used to analyze cyber-security databases for various kinds of cyber-attacks: specific, recall, f1-score, correctness, accuracy, and constancy. The two's respective precision, detection rate (DR), and totally bogus rate were also compared to see which one outperforms the other (FPR). The solution was then contrasted with the IDS that was already in place. In terms of detecting network anomalies, The outcomes show that PSO + ANN outperforms the PSO + KNN and PSO + DT classifier algorithms.

show abstract

Hybrid Particle Swarm Optimized Models for Rainfall Prediction: A Case Study in India

Zoremsanga,

Hussain

2024

Pure Appl. Geophys.

View full text Add to dashboard Cite

Sound classification using evolving ensemble models and Particle Swarm Optimization

Cited by 29 publications

References 83 publications

A Deep Ensemble Neural Network with Attention Mechanisms for Lung Abnormality Classification Using Audio Inputs

A Deep Ensemble Neural Network with Attention Mechanisms for Lung Abnormality Classification Using Audio Inputs

Mechanism for Extracting Features Using Particle Swarm Optimization for Intrusion Detection Systems

Hybrid Particle Swarm Optimized Models for Rainfall Prediction: A Case Study in India

Contact Info

Product

Resources

About