Machine learning, waveform preprocessing and feature extraction methods for classification of acoustic startle waveforms

Abstract

The acoustic startle response (ASR) is an involuntary muscle reflex that occurs in response to a transient loud sound and is a highly-utilized method of assessing hearing status in animal models. Currently, a high level of variability exists in the recording and interpretation of ASRs due to the lack of standardization for collecting and analyzing these measures. An ensembled machine learning model was trained to predict whether an ASR waveform is a startle or non-startle using highly-predictive features extracted from normalized ASR waveforms collected from young adult CBA/CaJ mice. Features were extracted from the normalized waveform as well as the power spectral density estimates and continuous wavelet transforms of the normalized waveform. Machine learning models utilizing methods from different families of algorithms were individually trained and then ensembled together, resulting in an extremely robust model. ASR waveforms were normalized using the mean and standard deviation computed before the startle elicitor was presented 9 machine learning algorithms from 4 different families of algorithms were individually trained using features extracted from the normalized ASR waveforms Trained machine learning models were ensembled to produce an extremely robust classifier

College English Reading Teaching Integrating Production Oriented Approach from the Perspective of Artificial Intelligence

“…N represents the total number of students, and n l means the number of students with behavioral characteristic l. For the test data of the numerical class, the original data is scaled by a certain range mapping through normalized preprocessing [30].…”

Investigation of data augmentation techniques for ultrasonic acoustic fiber sensing signals in guided wave-based pipeline damage detection

“…Such variability often mirrors the natural fluctuations found in experimental data. In Timothy's study [38], histogram analysis was employed to distinguish startle from non-startle ASR waveforms using crucial features. The histograms showed how features like maximum magnitude and their timing varied, with startle responses typically displaying greater magnitudes compared to non-startle responses.…”

Design of automatic identification algorithm for double-feature fault signal waveform of power equipment

“…Most distribution equipment can obtain fault signals through neutral grounding, but this process is difficult and susceptible to zero-sequence current, so an effective dual-feature fault signal waveform identification algorithm for power equipment must be used as support [18][19][20] . Relevant researchers have designed several dual-feature fault signal waveform automatic identification algorithms for power equipment according to the operating characteristics of power equipment and the requirements for automatic identification of fault signal waveforms for power equipment.…”