Optimal Mother Wavelet Selection for Machine Learning Signal Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for selecting mother wavelets for wavelet transforms in machine learning tasks lack the ability to identify optimal wavelets, affecting the accuracy of signal analysis and feature extraction.
Innovation Solution
A method and system that compute energy and entropy values, calculate distance between centroids, and normalize distance values using the N-norm technique to identify optimal mother wavelets for wavelet transforms, facilitating better signal classification and regression in machine learning tasks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional wavelet selection methods (CWT) are used, then wavelet analysis can be performed, but the selection of optimal mother wavelet for machine learning tasks is not achieved
Solution Approach 1:
The patent changes the selection criterion from traditional CWT-based methods to a machine learning-oriented parameter set including energy, entropy, centroid distance, and standard deviation. These parameters are computed for each candidate mother wavelet to objectively evaluate their suitability for classification and regression tasks, thereby improving signal analysis accuracy while providing a systematic selection framework.
Solution Approach 2:
The patent replaces the traditional mechanical/manual wavelet selection process with an automated computational system. The system automatically computes multiple parameters for each candidate wavelet, compares them against training data, and selects the optimal wavelet based on performance metrics, eliminating the need for manual trial-and-error selection and reducing complexity.
2Reliability
If multiple parameters (energy, entropy, centroid distance, standard deviation) are computed for wavelet selection, then optimal mother wavelet identification is improved, but computational complexity increases
Solution Approach 1:
The patent performs preliminary computations of energy, entropy, centroid distance, and standard deviation for all candidate mother wavelets before the actual selection process. By pre-computing these parameters and storing them, the system avoids redundant calculations during the selection phase, thereby improving selection reliability while reducing the computational power required at runtime.
Solution Approach 2:
The patent computes a set of parameters that is sufficient (but not excessive) for wavelet selection. Rather than computing all possible wavelet characteristics, it focuses on the four most relevant parameters (energy, entropy, centroid distance, standard deviation) that have been shown to effectively distinguish optimal wavelets for machine learning tasks, balancing reliability with computational efficiency.
3Measurement precision
If distance between centroids and standard deviation are used as selection criteria, then classification performance is improved, but the complexity of the selection process increases
Solution Approach 1:
The patent creates a universal selection framework that uses centroid distance and standard deviation as multi-functional criteria applicable to both classification and regression tasks. These parameters serve multiple purposes: they measure class separation, evaluate wavelet suitability, and guide the selection process, thereby improving signal class distinction accuracy while maintaining a unified, easy-to-operate selection mechanism.
Solution Approach 2:
The patent enables the wavelet selection process to be self-service by automatically computing centroid distances and standard deviations, comparing them against training data, and selecting the optimal wavelet without manual intervention. The system self-evaluates candidate wavelets based on these parameters and makes the selection autonomously, improving both accuracy and ease of operation.
Data Source
AI summary
Systems and methods for obtaining optimal mother wavelets for facilitating machine learning tasks. The traditional systems and methods provide for selecting a mother wavelet and signal classification using some traditional techniques and methods but none them provide for selecting an optimal mother wavelet to facilitate machine learning tasks. Embodiments of the present disclosure provide for obtaining an optimal mother wavelet to facilitate machine learning tasks by computing values of energy and entropy based upon labelled datasets and a probable set of mother wavelets, computing values of centroids and standard deviations based upon the values of energy and entropy, computing a set of distance values and normalizing the set of distance values and obtaining the optimal mother wavelet based upon the set of distance values for performing a wavelet transform and further facilitating machine learning tasks by classifying or regressing, a new set of signal classes, corresponding to a new set of signals.