Home Library Explore SAT

Home Library Explore SAT

comm lec 20 Flashcards | StudBud

Explore/comm lec 20

Why should we care about learning probability and statistics for communications?

Probabilistic models are better than deterministic models because (Note: ML = maximum likelihood method)

comm lec 20

4 terms·StudBud Administration·5 plays

Imported from external platform

Flashcards4 cardsFPractice quizPHost liveL

Questions 4

Answers

1
Probabilistic models are better than deterministic models because (Note: ML = maximum likelihood method)
None of the above
they allow for sophisticated machine learning techniques to solve problems
they serve as a conduit for statistical techniques, such as ML, to be used
they can account for uncertainty in the signals and/or systems
2
Why should we care about learning probability and statistics for communications?
Transmitter design often requires consideration of random nature of signals
Recovery of transmitted signals often employs statistical techniques
Ease the analysis of communication systems
Additive noise is random
3
Maximizing the posterior probability for transmit signal recovery
must have accurate prior probability to perform better than ML
is the best Bayesian method for doing that
always have better performance vs. ML as it fuses likelihood and prior prob
is not a good idea as it requires prior knowledge of the transmit signal
4
Parametric distribution (ppdf) is better than nonparametric (npdf) because
finding the parameters of a ppdf is easier than finding the npdf
easier to do analysis vs. nonparametric
some allows us to compute probability of certain events easily
all random phenomenons can be modeled using parametric distribution