Machine Learning for Environmental Noise Classification in Smart Cities

Ali Othman Albaji (Autor)

Buch | Hardcover

XV, 206 Seiten

2024 | 1st ed. 2024
Springer International Publishing (Verlag)
978-3-031-54666-2 (ISBN)

Artikel merken

We present a Machine Learning (ML) approach to monitoring and classifying noise pollution. Both methods of monitoring and classification have been proven successful. MATLAB and Python code was generated to monitor all types of noise pollution from the collected data, while ML was trained to classify these data. ML algorithms showed promising performance in monitoring the different sound classes such as highways, railways, trains and birds, airports and many more. It is observed that all the data obtained by both methods can be used to control noise pollution levels and for data analytics. They can help decision making and policy making by stakeholders such as municipalities, housing authorities and urban planners in smart cities. The findings indicate that ML can be used effectively in monitoring and measurement. Improvements can be obtained by enhancing the data collection methods. The intention is to develop more ML platforms from which to construct a less noisy. The second objective of this study was to visualize and analyze the data of 18 types of noise pollution that have been collected from 16 different locations in Malaysia. All the collected data were stored in Tableau software. Through the use of both qualitative and quantitative measurements, the data collected for this project was then combined to create a noise map database that can help smart cities make informed decisions.

Ali Othman Albaji received a bachelor's degree in electrical engineering specializing in "General communications" from the Civil Aviation Higher College, Tripoli, Libya, in 2007, and a Master's degree in electronics and telecommunication engineering from University Technology Malaysia *UTM*, Johor Bahru, Malaysia in 2022. His research interests are Machine Learning (ML), IoT, Wireless Sensor Networks (WSN), VSAT, SCADA Systems, Optical Networking, Wireless Communications, Deep Learning (DL), Artificial intelligence (AI), Web design, Robotics, and Programming Languages expert / Traineron ( Python, MATLAB, JAVA, JAVA Script, SQL, Data Base MSQL, C++, HTML, and....ETC).

CHAPTER 1 INTRODUCTION ... 1
1.1 Overview ... 1
1.2 Problem Statement ... 3
1.3 Research Objectives ... 4
1.4 Scope of Project ... 4
1.5 Thesis Outline ... 5

CHAPTER 2 LITERATURE REVIEW ... 7
2.1 Introduction ... 7
2.2 Research Background ... 9
2.3 Data analytics and data visualization dashboard ... 15
2.4 Machine Learning ... 15
2.4.1 Supervised learning ... 16
2.4.2 Unsupervised learning ... 17
2.5 Machine Learning Algorithms ... 17
2.5.1 Decision Tree (DT) ... 17
2.5.2 Logistic Regression (LR) ... 18
2.5.3 K-nearest-neighbor (KNN) ... 19
2.5.4 Support Vector Machine (SVM) ... 20
2.5.5 Random Forest (RF) ... 21
2.6 Machine Learning parameters ... 22
2.6.1 Confusion Matrix ... 22
2.6.2 Classification Accuracy ... 23
2.6.3 Precision ... 23
2.6.4 Recall ... 23
2.6.5 F1-Score ... 23
2.7 MATLAB software ... 24
2.8 Python software ... 24
2.9 Tableau software ... 25
2.10 Effects of Noise Pollution ... 25
2.10.1 Effects of Noise on Older Adults ... 26
2.11 Perceptions of Noise ... 26
2.12 Fundamentals of Noise ... 28
2.12.1 Individual Vehicles ... 28
2.12.2 Aircraft Noise ... 29
2.12.3 Wind-Turbine Noise ... 29
2.12.4 Mechanical Noise ... 30
2.12.5 Railway Noise ... 30
2.13 Environmental Noise Modeling and Monitoring ... 31
2.14 Conservation Program and Control Measures ... 34
2.15 Existing Apps for Noise Data Capturing ... 36
2.15.1 NoiseCapture App ... 36
2.15.2 Too Noise Pro ... 36
2.15.3 NoisePlatform ... 37
2.16 Weighting Filters in Noise Measurements ... 37
2.16.1 Frequency Weighting ... 38
2.16.2 Time Weighting ... 38
2.17 Previous Works ... 40
2.18 Summary ... 54

CHAPTER 3 RESEARCH METHODOLOGY ... 55
3.1 Introduction ... 55
3.2 Project Flowchart ... 55
3.3 Proposed Machine Learning Based Approach for Noise Classification ... 56
3.4 Qualitative Data ... 59
3.4.1 Qualitative analysis based on a survey using SPSS ... 60
3.5 Development of an Interactive Web Dashboard ... 62
3.6 Summary ... 62

CHAPTER 4 RESULTS AND DISCUSSION ... 63
4.1 Introduction ... 63
4.2 Machine Learning classification using MATLAB ... 63
4.2.1 Validation Receiver Operating Characteristic curve
(ROC Curve) ... 64
4.2.2 Parallel Coordinates ... 66
4.2.3 Validation Confusion Matrix ... 67
4.3 Machine Learning classifications using Python ... 68
4.3.1 Introduction ... 68
4.3.2 Data preparation ... 68
4.3.3 Data analysis ... 69
4.3.4 Noise samples captured over Malaysian cities ... 70
4.3.5 Classification Models ... 73
4.3.6 Confusion Matrix ... 77
4.3.7 Relation between cities and noise types ... 78
4.4 Benchmark between (MonitorNoises Vs Yaseen et.al [14]) ... 85
4.4.1 Performance Comparison with Benchmarking ... 86
4.4.2 MonitorNoises (2023) ... 88
4.4.3 Project Improvement ... 90
4.5 Data warehousing using Tableau software ... 93
4.6 A Survey based on Noise Pollution Monitoring ... 99
4.6.1 Introduction ... 99
4.6.2 Relationship Knowledge of Noise Pollution with
Demographic Variable ... 104
4.6.3 Perception of Respondents regarding Noise Pollution ... 105
4.6.4 Correlation Analysis ... 112
4.7 Chapter Summary ... 114

CHAPTER 5 CONCLUSION AND RECOMMENDATIONS ... 117
5.1 Introduction ... 117
5.2 Contributions to Knowledge (REWRITE) ... 118
5.3 Future Works ... 119

REFERENCES 121

Erscheinungsdatum	23.03.2024
Reihe/Serie	Synthesis Lectures on Engineering, Science, and Technology
Zusatzinfo	XV, 206 p. 44 illus., 39 illus. in color.
Verlagsort	Cham
Sprache	englisch
Maße	168 x 240 mm
Themenwelt	Technik ► Maschinenbau
Schlagworte	Artificial Intelligence • machine learning • MATLAB • Noise Pollution • Python • smart cities • urbanity
ISBN-10	3-031-54666-0 / 3031546660
ISBN-13	978-3-031-54666-2 / 9783031546662
Zustand	Neuware