classical probability
uses theory to calculate probability (math)
empirical probability
only uses observation to calculate probabilities (stats)
Subjective Probability
uses personal knowledge to create likelihoods (decision making ability)
The Law of Large Numbers
as experiments are repeated without bound the observed probability will approach the theoretical probability
probability
science of attaching likelihoods to random occurrences
experiment
any process with the likelihood of random outcomes
sample space
the set of all possible outcomes of an experiment
the event
subset of the sample space, what we want to see/successful outcomes
Basic probability formula
probability of event = #outcomes in event/#outcomes in sample space (P(E)=E/S)
mutually exclusive
when 2 events have no outcomes in common
basic addition rule
if A&B are mutually exclusive then probability of A OR B is P(A) + P(B)
the addition rule
when not mutually exclusive P(A&B) = P(A)+ P(B) - P(A&B)
complements
an event E is the complement of E^c if they have no outcomes in common and E+E^c equals the total sample space
the multiplication rule
for when events happen in succession:
events must be independent
P(A&B)= P(A)* P(B) “and then”
conditional probability
P(A|B) “A given B”
this uses a given condition that causes events to have a dependency
geometric probability
used for probabilities looking for the first success
formula: P(nth) = pq ^n-1
p-success
q-failure
binomial probability
-need a set # of trials
-only two outcomes (success/failure)
-all trials are independent
P( exactly A) =p^A * q^Ac * nCr
p-probability success; q-probability failure; Ac- complement of outcomes/ n-A; nCr-combination
probability distributions
a table that represents the outcomes of an experiment with the given probabilities
x | P(x) | xP(x)
the mean of xP(x) is Expected Value
EV = sum of xP(x)
Normal Distribution
mean falls in the middle
graph is symmetrical
graph floats above x axis
area under whole graph = 1
bell shaped curve
Standard Normal Distribution
zero fall in the middle
measure by std dev
graph is symmetrical
graph floats above x axis
area under curve = 1
z-scores
measure the distance from the mean using std devs
z = (x-mean)/std dev
percentiles
always lean left tail/shading
75th percentile means less than 75%
central limit theorem
when data ISNT normal
as the size of samples taken from a population increases without bound the graph of the sample means will approach normal and the mean of the sample means will equal the population mean and the std dev of the sample means will equal the population std dev divided by the square root of the sample size
n>30 assume normal
mew of x-bar = mew
std dev of x-bar = std dev/root(n)
new z-score formula
z = (xbar-mew)/(sigma/root(n))
proportions
p-hat - sample proportion for success
w-hat - sample proportion for failure
central limit theorem for proportions
if nqp > 10 then assume normal
z = (phat-p) / root(pq/n)
binomial probability
“at most” “at least”
use binomial calculator