01:830:305 Lecture Notes - Lecture 22: False Positives And False Negatives, Conditional Probability, Daniel Kahneman
20 Apr 2016
School
Department
Course
Professor
Document Summary
Ex: you think you grandma walks into this room bc it looks just like her then think wait my grandma is 1000 miles away, not possible or likely denominator does really matter because it accounts for all the possibilities and is the same for every formula. all we care about is product of product and likeihood. is it your grandma? given the data how strongly you believe it is time how plausible it is. Famous medical example just have a screening, no symptoms before and you test positive for the disease. there is a difference between testing positive and actually being sick. usually correct but not always highly sensitive, test is actually a very good test. if test positive . 99% you have the disease. missed healthy people p (+ld) highly specific, healthy people test negative. 95% (5% of positive tests are false positives) > healthy sick people p( l~d)
