06 - Probabilistic models and logistic regression


  • Description:: models and regression

We can use two models:

discriminative: given one input say what’s the probability to classify it in one class (first estimate and the compute using Bayes

or we can use a generative approach: estimate class condition density directly

P(C1) = Pi1 P(C2) = Pi2 … P(Ck-1) = Pi k-1 P(Ck) = 1 - sum of all of them

for 8 class: 8 params but you must reply k-1


There is a parameter called pi that is the sum between positive and negative over number of things we have. pi = 1 + 1 / N