Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
33 Cards in this Set
- Front
- Back
asymptotic efficiency
|
use model that has closest forecast error to the true model
|
|
AIC
|
assymptotically efficient
|
|
SIC
|
consistent
|
|
testing down
|
general to specific
|
|
unstructured data
|
cross sectional
|
|
structured data
|
time series
|
|
seasonality
|
a regular deviation from the overall average that is tied to the periodicity of the data
|
|
multiplicative seasonality
|
seasonality is a percentage of base trend- can take log
|
|
additive seasonality
|
seasonality is a fixed level above or below base trend
|
|
variance=
|
sigma squared
|
|
k=
|
number of right hand side variables
|
|
anscombe's quartet
|
four datasets with identical statistics, but very different graphs
|
|
aspect ratio
|
ratio of graph's height to its width- should equal .618
|
|
banking to 45 degrees
|
select aspect ratio to make the average absolute slope of the line segments equal to 45 degrees
|
|
bivariate scatterplots
|
graph of y versus x, for example
|
|
chartjunk
|
elaborate shadings and grids, decoration, and related nonsense
|
|
exploratory data analysis
|
learning about patterns in data without pretending to have too much a priori theory
|
|
golden ratio
|
aspect ratio of .618
|
|
histogram
|
provides a simple estimate of the probability density of a random variable
|
|
innovation outlier
|
driven by shocks whose effects naturally last more than one period
|
|
measurement outlier
|
reflect measurement errors
|
|
multiple comparisons
|
graphing a number of datasets in one figure
|
|
multiway scatterplot
|
scatterplot matrix- the set of all possible bivariate scatterplots
|
|
nondata ink
|
ink used to depict anything other than data points
|
|
pairwise scatterplot
|
=bivariate scatterplot
|
|
partial correlation
|
examine the correlation between x and y controlling for z
|
|
relational graphics
|
used to display relationships and flag anomalous observations
|
|
simple correlation
|
studies relationship between say, x and y
|
|
time series plot
|
series is graphed against time
|
|
exponential trend
|
log linear trend
|
|
in sample overfitting/data mining
|
including more variables in a model won't always improve its out of sample forecasting performance
|
|
SIC
|
higher penalty for higher degrees of freedom
|
|
consistency
|
higher probability of selecting the right model
|