universal estimator exp-2

yossi-cohen commented 3 years ago

Show that the error of predicting the input data is "same" as for predicting the test-set Calculate the variance of the error produced in #11

var_1 = E [ ( e - mean(e) ) ^ 2 ] = 1/n * sum ( e_i - mean(e) ) ^ 2 sigma_1 = sqrt ( var_1 )

Calculate the avg. error of sigma_1:

var_2 = 1/n * sum [ ( e_i - mean(e) ) ^ 2 - sigma_1 ^ 2 ] sigma_2 = sqrt ( var_2 )

Plot a graph:

yossigil commented 3 years ago

please formulate a research question

yossigil commented 3 years ago

????

yossi-cohen / preferential-attachment