Non-sinusoidal waveform model

Click For Summary
SUMMARY

The discussion centers on a ninth-grade student's attempt to create a regression algorithm in Desmos to model non-sinusoidal waveforms for predicting binary inputs. The student has experimented with Fourier's theorem but encountered issues with noise as more parameters are added. A participant pointed out that the current model is linear near the origin, suggesting that the parameters used do not effectively capture the desired non-sinusoidal behavior. The conversation highlights the challenges of predicting random binary sequences and the limitations of linear models in this context.

PREREQUISITES
  • Understanding of regression algorithms
  • Familiarity with Fourier's theorem
  • Basic knowledge of waveforms and their properties
  • Experience using Desmos for mathematical modeling
NEXT STEPS
  • Research non-linear regression techniques for modeling waveforms
  • Explore advanced Fourier analysis methods for better fitting
  • Learn about polynomial regression in Desmos
  • Investigate methods for predicting sequences in random data
USEFUL FOR

Students, educators, and hobbyists interested in mathematical modeling, particularly those exploring regression algorithms and waveform analysis in Desmos.

P3dr0
Messages
1
Reaction score
0
Hi, I am a ninth-grade student from Portugal with nothing to do, and I have decided that I want to build a simple regression algorithm in Desmos (the online calculator) to fit random binary inputs and maybe predict the next binary digits (although that part may take a considerable amount of time). For now, I am just trying to find a way to create non-sinusoidal waves, but I can't find any model for it. Can anybody tell me a simple yet effective model for what I am trying to achieve?
PS: I have created an algorithm based on Fourier's theorem, but it doesn't work quite well. (I'll attach a photo and a link of what I have done.) This is the simplest version I have created so far, but I have tried other parameters. However, when I add more elements (more parameters), the function becomes noisier.

Link: https://www.desmos.com/calculator/h3nww18l8p
 

Attachments

  • Captura de ecrã 2024-04-03 190941.png
    Captura de ecrã 2024-04-03 190941.png
    11.9 KB · Views: 94
Mathematics news on Phys.org
What you have done is perfectly correct, but I am not sure that it is a good example of what you want to do. Notice that your example is perfectly linear with a slope of 1 and very near the (x=0, y=0) origin. The sin() function is also very linear with a slope of 1 near the origin. The tiny value of ##b## means that the inputs of sin() will be very close to 0 and the large value of ##a## means that the slope will be close to 1. If ##a=1/b##, the slope would be exactly 1 at the origin.
 
Welcome to PF.

P3dr0 said:
to fit random binary inputs and maybe predict the next binary digits
If the inputs are random, how can you do any predictions for future digits?
 

Similar threads

  • · Replies 23 ·
Replies
23
Views
4K
Replies
3
Views
3K
  • · Replies 8 ·
Replies
8
Views
3K
Replies
15
Views
3K
  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 8 ·
Replies
8
Views
2K
  • · Replies 6 ·
Replies
6
Views
945
  • · Replies 1 ·
Replies
1
Views
1K
  • · Replies 6 ·
Replies
6
Views
3K
  • · Replies 2 ·
Replies
2
Views
1K