I have a gaussian distributed time series ($X_t$) with some parameters in my experiment. Suppose I want to know the mean $\mu$. If I define another time series $Y_t$ such that $Y_t=X_t-a$ for all $t$. Now say I vary this parameter $a$ and generate altogether different time series for each $a$, say $Y_t(a)$. I look at the mean of $Y_t$ for each $a$. The value of a, where I get the mean of $Y_t$ closest to $0$, will be my estimate of $\mu$. Say I will eventually use this learnt value of $\mu$ to generate $Y_t$ as my final goal. Can this be called ML? I am using some training data of $X_t$ to learn about its parameter and then using test data of $X_t$ to generate $Y_t$.

Now why am I working so hard on this simple problem? Well, actually I am not. I am doing something else, which will have lots of parameters in the time series and will be used to generate other time series after similar parameter extraction. That will be too complicated to discuss here. I just wanted to clear my basics using an over-simplified example.

  • $\begingroup$ I don't understand what you mean by "gaussian distributed time series". Can you precisely define it? Are you saying that each element of the time series is sampled from 1. the same or 2. possibly different Gaussian? In general, I don't fully understand your problem. Do you have the original time series $X_t$? If not, do you want to estimate it from "noisy" similar time series? I don't understand how subtracting $a$ from $X_t$ gives you any insight. Also, I don't understand how you are going to generate $Y_t$ if you don't have $X_t$. Are you assuming that $Y_t$ is $X_t -a$, for some $a$? $\endgroup$
    – nbro
    Apr 25, 2020 at 23:44
  • $\begingroup$ @nbro Thanks for your response. The time series $X_t$ is given, and we know it has a Gaussian distribution, (histogram: Gaussian). So each point in $X_t$ belongs to a gaussian distribution with the same mean and variance. Now if you subtract a from each point in $X_t$ to construct a different series $Y_t$ the mean of the gaussian shifts. For some $a$, the constructed $Y_t$ has 0 mean. What I am asking if you vary $a$ and try to see for what $a$ this is achieved, then use that correctly found a, to generate more data points of $Y_t$, from $X_t$ can I call it as ML? $\endgroup$
    – Raunak Dey
    Apr 27, 2020 at 17:40

1 Answer 1


Does it classify as Machine Learning? If I understand the problem correctly, my opinion is yes, this is machine learning. In fact, I think it is just linear regression for finding the equation of a horizontal line y(t) = a that fits the the dataset y(t) = Xt.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .