Signal Processing and Machine Learning for Innovation Engineering

This book presents recent advances in signal processing and artificial neural network (ANN) applications aimed at driving innovation in engineering. The proposed developments constitute advanced IT solutions for research, data access, and knowledge management, primarily built upon Internet of Things (IoT)-assisted architectures. The techniques described integrate signal processing with wireless neural network implementations. Several signal processing, feature selection, and feature extraction approaches are explored, including the use of bandpass filters combined with numerical derivatives, the Hilbert transform, adaptive thresholding, moving average filters, autocorrelation functions, wavelet transforms, and the Hilbert–Huang transform. For feature selection, methods such as feature normalization and the False Nearest Neighbor (FNN) technique are examined, while principal component analysis (PCA) is applied for feature extraction. The book also presents real-time tests to assess IoT-assisted architectures using acquired signals, focusing on the accuracy and performance of ANN models in various tasks such as prediction, modeling, control, monitoring, and classification. The research applications discussed encompass fault diagnosis, arrhythmia classification, electrocardiogram (ECG) analysis for global health monitoring, autoscaling systems, and anomaly detection. Overall, this work offers a comprehensive overview of how the integration of signal processing and ANN techniques within IoT frameworks can advance intelligent engineering systems.