The “excess” value of this quantity is defined by subtracting three from it, since that’s this quantity’s value for the normal distribution. For 10 points each:
[10m] Name this quantity computed by rescaling the expected value of a variable’s fourth power.
ANSWER: kurtosis [accept excess kurtosis]
[10e] Distributions with kurtosis greater than three have “heavy” examples of these features at the outer ends of a distribution, which represent the probability of extreme values. Distributions with fat examples of these features may have large numbers of outliers.
ANSWER: tails [accept heavy tails or fat tails]
[10h] This heavy-tailed distribution has a kurtosis exactly twice that of the normal distribution. This distribution, which has a cusp at its mean, is the prior distribution of the coefficients in the Bayesian LASSO method.
ANSWER: Laplace (“luh-PLOSS”) distribution [or double exponential distribution]
<Other Science>