Today's study group.   Example of the union bound.

X is a continuous random variable, thus
P(X = 3) = 0
P(X = n) = 0
and we do not have to weary about boundaries.

P(A1) = P[0 <=  X <= 2]  = 1/(2^2)
P(A2) = P[1 <=  X <= 3]  = 1/(2^3)
P(A3) = P[2 <=  X <= 4]  = 1/(2^4)
.............................

by the union bound:
P(union Ai)  <=  1/(2^2) + 1/(2^3) + 1/(2^4) +...

Recording of last ML study group - comparing to models -
https://www.youtube.com/watch?v=ytFJ4aTT9S0&list=PLC7m9qp0Q1Ye3Fwnooz2k5xh0p5eD1s4k&index=17

The union bound and its significance in learning.   https://drive.google.com/file/d/1V3fKepfnhj3EESeKfLbJsz-bnDu96sSD/view?usp=sharing

We compare two models using the confidence interval
https://colab.research.google.com/drive/1v6VudBYzgrymG-ZjQqMTWKlsvRm0GAUb?usp=sharing

A ML example of confidence interval of performance using bootstrapping https://www.youtube.com/watch?v=G5OgofpRnY4&list=PLC7m9qp0Q1Ye3Fwnooz2k5xh0p5eD1s4k&index=17

We'll revisit the bootstrapping example and cast it in the context of a ML learning example  (example 2) https://www.overleaf.com/read/kdvqgpkxnhjt

The last workshop on convergence in distribution is recorded here -
https://www.youtube.com/watch?v=dB8zdcUMAyI&list=PLC7m9qp0Q1Ye3Fwnooz2k5xh0p5eD1s4k&index=17&t=0s

We'll discuss convergence in distribution.  See https://www.overleaf.com/read/kdvqgpkxnhjt