Is there any difference between the Poisson process and the sigmoid (s-curve) function? If the traffic generation at each device can be modeled as a Poisson process, is it possible to exploit (sigmoid) to quantify the importance of the information carried by that data traffic?
You seem to be a little confused between logistic and poisson, sometimes called polychotomus logistic regression, if I remember my biostatistics correctly. This book may help:
https://www.amazon.com/Fundamentals-Biostatistics-Bernard-Rosner/dp/130526892X/ref=sr_1_1?s=books&ie=UTF8&qid=1550128079&sr=1-1&keywords=fundamentals+of+biostatistics+8th+edition
Look in the index. see also this link:https://ncss-wpengine.netdna-ssl.com/wp-content/themes/ncss/pdf/Procedures/NCSS/Poisson_Regression.pdf I apologize if I made an error here.
Best Wishes, David Booth
It really depends on the distribution of random variables that you are putting into the model. sometimes these function are called kernel in machine learning. if your kernel is gaussian distribution, your process will be gaussian process, if it is poisson distribution , your model will be poison process and if your kernel is gumbel distribution it will be logistic process.
- Badges
- Science topic