lecture 4: statistical inference
DESCRIPTION
Basic concepts of statistical inference. Outline: stochastic variables, frequency functions, expectations, variance, entropy, joint probabilities, conditional probabilities, independence, sampling, estimation, maximum likelihood estimation (MLE), smoothing, hypothesis testing,z-test,TRANSCRIPT
Machine Learning for Language Technology Lecture 4: Sta,s,cal Inference
Marina San,ni Department of Linguis,cs and Philology Uppsala University, Uppsala, Sweden
Autumn 2014
Acknowledgement: Thanks to Prof. Joakim Nivre for course design and materials
Stochas,c Variables
Types of Variables
Frequency Func,ons
Expecta,on
Variance
Entropy
More on Entropy
Joint and Condi,onal Probability
Independence
Part-‐of-‐Speech Bigrams 1
Part-‐of-‐Speech Bigrams 2
Part-‐of-‐Speech Bigrams 3
Sta,s,cal Inference
Sampling
Es,ma,on
Maximum Likelihood Es,ma,on (MLE)
MLE: Example 1
MLE: Example 2
MLE: Ra,onale
MLE and Smoothing
Interval Es,ma,on
More on Interval Es,ma,on
Hypothesis Tes,ng
Example: Z-‐test
The end