0

I am new about these topics. I research a lot of article about this issue. There are a lot of different techniques. But I am confused, because I don't know, where to start.

According to my research, first important thing; I must make preprocessing to the raw sensor data. There are some techniques, fft is one of them. (But how can I search to learn all techniques? I did not see all techniques in same page.)

Then I start the statistical calculates to processing.

I did not draw a roadmap. Can you help these issue or suggest books or anything?

özlem bulu
  • 31
  • 2
  • 5

2 Answers2

0

Welcome to SO ... to leverage this site hover your mouse over top of tag fft on your question ... then click View tag ... then hit learn more ... then after reading the info page on fft hit Votes to see the highest voted posts here on SO ... those questions/answers will get you into the ball park

I highly suggest you master the details explained here Discrete Fourier Transform - Simple Step by Step

An Interactive Guide To The Fourier Transform
https://betterexplained.com/articles/an-interactive-guide-to-the-fourier-transform/

Intuitive Understanding of the Fourier Transform and FFTs
https://www.youtube.com/watch?v=FjmwwDHT98c

An Intuitive Discrete Fourier Transform Tutorial
http://practicalcryptography.com/miscellaneous/machine-learning/intuitive-guide-discrete-fourier-transform/

How to get frequency from fft result?

I could go on mentioning nuggets from my notes however I will leave you with this excerpt from an excellent book

http://www.dspguide.com/ch10/6.htm

The Discrete Time Fourier Transform (DTFT) is the member of the Fourier transform family that operates on aperiodic,
discrete signals. The best way to understand the DTFT is how it relates to the DFT. To start, imagine that you
acquire an N sample signal, and want to find its frequency spectrum. By using the DFT, the signal can be
decomposed into sine and cosine waves, with frequencies equally spaced between zero and one-half of the
sampling rate. As discussed in the last chapter, padding the time domain signal with zeros makes the period
of the time domain longer, as well as making the spacing between samples in the frequency domain narrower.
As N approaches infinity, the time domain becomes aperiodic, and the frequency domain becomes a continuous signal.
This is the DTFT, the Fourier transform that relates an aperiodic, discrete signal, with a periodic,
continuous frequency spectrum
Scott Stensland
  • 26,870
  • 12
  • 93
  • 104
0

The first step will be data cleaning and feature extraction. You need to prepare data in format that is applicable to Machine Learning algorithms. I recommend to you my paper "Generic Data Imputation and Feature Extraction for Signals from Multifunctional Printers". It is about preparing data from IoT signals for further application of ML algorithms.

wind
  • 892
  • 1
  • 11
  • 27