Very well explained. I have one question. If autocorrelation function is maximum at amplitude 2 and one tells no correlation, then how come autocorrelation at time t (at same time) is one?
The correlation function when there's no correlation should be one. So, at long times (if you wait long enough) the signal is not correlated with itself, so the convergence value of the correlation function will be one. This amplitude is 1 (or 2 in case of autocorrelation) is just because he normalized the data for the sake of presentation.
In this video you defined the correlation as being the property that E[AB] isn't equal to E[A]*E[B] and hence you define the g function. Why don't you define the correlation between 2 rv using pearsons correlation definition? meaning cov(x,y) divises by the product of their stdev.
Very helpful lecture!
nice explanation
nice sir
Thank you very much sir
Very well explained. I have one question. If autocorrelation function is maximum at amplitude 2 and one tells no correlation, then how come autocorrelation at time t (at same time) is one?
The correlation function when there's no correlation should be one. So, at long times (if you wait long enough) the signal is not correlated with itself, so the convergence value of the correlation function will be one. This amplitude is 1 (or 2 in case of autocorrelation) is just because he normalized the data for the sake of presentation.
Thanks!
Is it possible to get the slides ?
In this video you defined the correlation as being the property that E[AB] isn't equal to E[A]*E[B] and hence you define the g function.
Why don't you define the correlation between 2 rv using pearsons correlation definition? meaning cov(x,y) divises by the product of their stdev.
Smith Timothy Jackson Scott Rodriguez William