Michaelmas Flashcards
(32 cards)
What is the difference between stochastic and deterministic?
Deterministic- one outcome for a given set of conditions
Stochastic- different outcomes when run, with the same conditions
What is a sample space?
Set of all possible outcome
What is a disjoint/ mutually exclusive event?
Not related, separate
P(AuB)= P(A) + P(B)
What are independent events?
Two events that don’t effect each other
What is conditional dependence?
When two things aren’t related but can be linked by an event
What is sensitivity in Bayes Theorem?
True positive, test positive and have disease
What is specificity in Bayes Theorem?
True negative, test negative and doesn’t have disease
How to calculate false positive?
1- specificity
What is typically the prior knowledge in Bayes Theorem?
The avg % in a population
What is the difference in permutation and combination?
Permutation- order matters
Combination- order doesn’t matter
What are the properties of a probability mas function (pmf)?
1) No probability of an outcome greater than 1
2) Sum of all probabilities equal 1
What is uniform distribution?
Probability of outcomes all equal
What is the Bernoulli trials?
When there 2 outcomes
When should you use the Binomial distribution?
- fixed number of trials
- probability doesn’t change
- binary outcome
When do you use hypergeometric geometric?
Determine probability of a CERTAIN number of successes without replacement
When should you use geometric distribution?
When want to find out how many times until first success
- probability consistent
When should you use poisson distribution?
Used to count numbers in a specific time/ distance etc.
What are the assumptions made in poisson?
- occur independently and randomly
- probability doesn’t change
- two events don’t occur at the same time
What is theta in poisson equal to?
Theta= lamda x time
How many occurrences in specified time
What does cdf stand for?
cumulative distribution function
When would you use pdf and pmf?
pdf= continuous
pmf= discrete
What are the 3 requirements for pdf to be true?
1) f(x)>0
2) Integration from -ve infinity to infinity =1
3) P(X=a) = 0 for all a
How to get from cdf (F(x)) to pdf (f(x)) ?
differentiate
How much of the graph is incorporated in 1 deviation and 2 deviations?
1 => 68.3%
2 => 95.4%
