Method and system for recommending features for developing an iot application
First Claim
1. A method for recommending a set of features for developing an Internet of Things (IoT) analytics application, the method comprising a processor implemented steps of:
- providing an input signal received from a sensor to the processor;
applying a four level discrete wavelet transform by selecting a suitable mother wavelet on the input signal to generate a time frequency domain (TFD) set of features;
applying a short term Fourier transform on the input signal to generate a frequency domain (FD) set of features;
generating a time domain (TD) set of features from the input signal, wherein the TFD set of features, the FD set of features and the TD set of features are an initial set of features;
applying a first feature selection algorithm and a second feature selection algorithm to the time domain, the frequency domain and the time-frequency domain set of features, wherein the application of selection algorithms result in selection of a first set of features for the time domain, the frequency domain and the time-frequency domain;
taking a union of the first set of features for TD, FD and TFD recommended by the first selection algorithm and the second selection algorithm;
generating a combination of the first set of features using an exhaustive search, wherein the combination of the first set of features is lesser than the initial set of features;
applying a classification algorithm including a support vector machine learning method using a different set of kernels on the combination of the first set of features; and
recommending the first set of features as the set of features for developing the IoT analytics application if a predefined condition is satisfied.
1 Assignment
0 Petitions
Accused Products
Abstract
A method and system has been provided for recommending features for developing an IoT analytics application. The method follows a deep like architecture. It comprises of three distinct layers. First layer is for input signal processing and other two layers are for feature reduction. The time domain, frequency domain and time-frequency domain features are extracted from the input signal. The invention uses multiple feature selection methods so that the union of the recommended features by these feature selection methods is significantly lesser than the initial set of features. The best feature combination is recommended using an exhaustive search.
9 Citations
12 Claims
-
1. A method for recommending a set of features for developing an Internet of Things (IoT) analytics application, the method comprising a processor implemented steps of:
-
providing an input signal received from a sensor to the processor; applying a four level discrete wavelet transform by selecting a suitable mother wavelet on the input signal to generate a time frequency domain (TFD) set of features; applying a short term Fourier transform on the input signal to generate a frequency domain (FD) set of features; generating a time domain (TD) set of features from the input signal, wherein the TFD set of features, the FD set of features and the TD set of features are an initial set of features; applying a first feature selection algorithm and a second feature selection algorithm to the time domain, the frequency domain and the time-frequency domain set of features, wherein the application of selection algorithms result in selection of a first set of features for the time domain, the frequency domain and the time-frequency domain; taking a union of the first set of features for TD, FD and TFD recommended by the first selection algorithm and the second selection algorithm; generating a combination of the first set of features using an exhaustive search, wherein the combination of the first set of features is lesser than the initial set of features; applying a classification algorithm including a support vector machine learning method using a different set of kernels on the combination of the first set of features; and recommending the first set of features as the set of features for developing the IoT analytics application if a predefined condition is satisfied. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10)
-
-
11. A system for recommending a set of features for developing an Internet of Things (IoT) analytics application, the system comprising:
-
a sensor for capturing an input signal; a preprocessor for preprocessing the input signal to remove the noise from the input signal; a memory; and a processor in communication with the memory, wherein the processor further comprising; a discrete wavelet transformation (DWT) module for applying a four level discrete wavelet transform by selecting a suitable mother wavelet on the input signal to generate a time frequency domain (TFD) set of features; a Fourier transformation module for applying a short term Fourier transform on the input signal to generate a frequency domain (FD) set of features; a time domain feature generation module generating a time domain (TD) set of features from the input signal, wherein the TFD set of features, the FD set of features and the TD set of features are an initial set of features; a first feature selection module for applying a first feature selection algorithm to the time domain, the frequency domain and the time-frequency domain set of features; a second feature selection module for applying a second feature selection algorithm to the time domain, the frequency domain and the time-frequency domain set of features, wherein the application of selection algorithms result in selection of a first set of features for the time domain, the frequency domain and the time-frequency domain; a union module for taking a union of the first set of features for TD, FD and TFD recommended by the first selection algorithm and the second selection algorithm; an exhaustive search module for generating a combination of the first set of features using an exhaustive search, wherein the combination of the first set of features is lesser than the set of features; a classification module for applying a classification algorithm including a support vector machine learning method using a different set of kernels on the combination of the first set of features; and a recommendation module for recommending the first set of features as the set of features for developing the IoT analytics application if a predefined condition is satisfied.
-
-
12. A non-transitory computer-readable medium having embodied thereon a computer program for developing an Internet of Things (IoT) analytics application, the method comprising a processor implemented steps of:
-
providing an input signal received from a sensor to the processor; applying a four level discrete wavelet transform by selecting a suitable mother wavelet on the input signal to generate a time frequency domain (TFD) set of features; applying a short term Fourier transform on the input signal to generate a frequency domain (FD) set of features; generating a time domain (TD) set of features from the input signal, wherein the TFD set of features, the FD set of features and the TD set of features are an initial set of features; applying a first feature selection algorithm and a second feature selection algorithm to the time domain, the frequency domain and the time-frequency domain set of features, wherein the application of selection algorithms result in selection of a first set of features for the time domain, the frequency domain and the time-frequency domain; taking a union of the first set of features for TD, FD and TFD recommended by the first selection algorithm and the second selection algorithm; generating a combination of the first set of features using an exhaustive search, wherein the combination of the first set of features is lesser than the initial set of features; applying a classification algorithm including a support vector machine learning method using a different set of kernels on the combination of the first set of features; and recommending the first set of features as the set of features for developing the IoT analytics application if a predefined condition is satisfied.
-
Specification