ISYE 6501 MIDTERM QUESTIONS &
ANSWERS(RATED A+)
What does SVM stand for? - ANSWERSupport Vector Machine
Is written text structured or unstructured? - ANSWERUnstructured
When we increase the sum of the square of the coefficients we... -
ANSWERDecrease the distance between the lines
In SVM soft classifier we tradeoff between maximizing ___ and minimizing ___ -
ANSWERmargin and errors
If lambda gets small what gets emphasized, large margin or minimizing training
error?, - ANSWERMinimizing errors.
What is a support vector? - ANSWERA point that holds up a shape.
Does ...[⅔(a-1)+1/3(a+1)] move an SVM classifier up or down? - ANSWERUp
How do you make errors more costly in a soft SVM classifier? - ANSWERinclude a
multiplier for the point-error term.
If an SVM coefficient is very close to zero... - ANSWERthat term is not very
important to the classification.
What is the difference between standardization and scaling? - ANSWERScaling is
bounded in range. Standardization is scaling to a normal distribution. Standardization
is the (value - factor mean) / (factor standard deviation)
What is the 2-norm? - ANSWEREuclidean distance
What is the 1-norm? - ANSWERThe rectilinear (Manhattan) distance
What is the infinity norm? - ANSWERThe value of the largest dimension
Measuring the quality of a model is called? - ANSWERValidation
What does a confusion matrix show? - ANSWERThe performance of a classification
model.
A time series outlier that seems "off the curve" is called a... - ANSWERcontextual
outlier.
A data element that is different from all other data in a set is called a... -
ANSWERpoint outlier.
,When something is missing in a range of points - ANSWERit is called a..., collective
outlier.
The whiskers on a box plot extend to... - ANSWERthe 10th and 90th percentiles (or
5th and 95th)
Why are hypothesis tests generally not sufficient for change detection? -
ANSWERThey are slow to detect changes.
In CUSUM, T is _____ and C is _____., - ANSWERThreshold and a "bring down
factor"
In a CUSUM model, you adjust T and C to manage the tradeoff between..., -
ANSWERearly detection and false-alarms
In exponential smoothing, if the data is less random, then you want to pick an alpha
that is..., - ANSWERClose to 1.
What is the initial condition for T in exponential smoothing with trending? -
ANSWERT_i=0
In cyclic exponential smoothing, L represents..., - ANSWERThe length of the cycle or
season
In cyclic exponential smoothing, C_1 ... C_L = ___?, - ANSWER1. In other words,
initialize it to no initial cycle.
Exponential, trending and cyclic smoothing are also referred to as - ANSWERsingle
double and triple.
Triple smoothing is also known as? - ANSWERWinter's or Holt-Winter's
What is the optimization formula for Exponential Smoothing? - ANSWERmin(F_t-
Xt)^2 where alpha and beta are between 0 and 1.
ARIMA stands for? - ANSWERAutoregressive Integrated Moving Average
Exponential smoothing is an order ___ autoregressive model. - ANSWERInfinity. It
uses data going all the way back.
For ARIMA, the D parameter is used to specify ___. - ANSWER, The order, or the
differences of the differences of the differences (d-times.)
For ARIMA, the P parameters is used to specify ____., - ANSWERThe order of
periods (autoregression).
For ARIMA, the Q parameter is used to specify ______., - ANSWERThe order of the
moving average.
ARIMA(0,1,1) is ?, - ANSWERExponential smoothing.
, What is the order of the ARIMA parameters? - ANSWERp d q
GARCH estimates what? - ANSWERVariance.
Variance can be a proxy for ___ or ___. - ANSWERvolatility or Risk
What parameter does GARCH not have the ARIMA has? - ANSWERd because
GARCH doesn't deal with differences.
What is a simple linear regression? - ANSWERLinear regression with one predictor.
A linear regression defines a relationship between what and what? -
ANSWERPredictor(S) and Response
What is the formula for a linear regression with M predictors? -
ANSWERY=a0+sum(j=1..m, a(j)*(x(j))
What measure is used to determine the quality of a linear regression line to data? -
ANSWERSquare of the difference between the line and the data points. (Sum
squared error.)
What is the formula for point error in linear regression? - ANSWERy(i)-Yhat(i)=y(i)-
(a(0) + a(i)*x(i)
Taking partial derivates and setting them equal to zero and then solving that system
equation helps us do what? - ANSWERMinimize the error and optimize the
coefficients for linear regression.
What does AIC stand for? - ANSWERAkaike Information Criterion
What is "Likelihood"? - ANSWERA measure for the probability density for any
parameter set.
What is Maximum likelihood? - ANSWERParameters that give the highest
probability.
What is MLE? - ANSWERMaximum Likelihood Estimate... The set of parameters
that minimizes the sum of square errors.
What is the formula for AIC? - ANSWERAIC=2k - 2*ln(L*) where L* is the maximum
likelihood value and K is the number of parameters estimated.
What is the penality term in AIC and what does it do? - ANSWER2K - It helps
prevent overfitting.
A models that is fit to random effects and not real ones is said to be? -
ANSWEROverfit
ANSWERS(RATED A+)
What does SVM stand for? - ANSWERSupport Vector Machine
Is written text structured or unstructured? - ANSWERUnstructured
When we increase the sum of the square of the coefficients we... -
ANSWERDecrease the distance between the lines
In SVM soft classifier we tradeoff between maximizing ___ and minimizing ___ -
ANSWERmargin and errors
If lambda gets small what gets emphasized, large margin or minimizing training
error?, - ANSWERMinimizing errors.
What is a support vector? - ANSWERA point that holds up a shape.
Does ...[⅔(a-1)+1/3(a+1)] move an SVM classifier up or down? - ANSWERUp
How do you make errors more costly in a soft SVM classifier? - ANSWERinclude a
multiplier for the point-error term.
If an SVM coefficient is very close to zero... - ANSWERthat term is not very
important to the classification.
What is the difference between standardization and scaling? - ANSWERScaling is
bounded in range. Standardization is scaling to a normal distribution. Standardization
is the (value - factor mean) / (factor standard deviation)
What is the 2-norm? - ANSWEREuclidean distance
What is the 1-norm? - ANSWERThe rectilinear (Manhattan) distance
What is the infinity norm? - ANSWERThe value of the largest dimension
Measuring the quality of a model is called? - ANSWERValidation
What does a confusion matrix show? - ANSWERThe performance of a classification
model.
A time series outlier that seems "off the curve" is called a... - ANSWERcontextual
outlier.
A data element that is different from all other data in a set is called a... -
ANSWERpoint outlier.
,When something is missing in a range of points - ANSWERit is called a..., collective
outlier.
The whiskers on a box plot extend to... - ANSWERthe 10th and 90th percentiles (or
5th and 95th)
Why are hypothesis tests generally not sufficient for change detection? -
ANSWERThey are slow to detect changes.
In CUSUM, T is _____ and C is _____., - ANSWERThreshold and a "bring down
factor"
In a CUSUM model, you adjust T and C to manage the tradeoff between..., -
ANSWERearly detection and false-alarms
In exponential smoothing, if the data is less random, then you want to pick an alpha
that is..., - ANSWERClose to 1.
What is the initial condition for T in exponential smoothing with trending? -
ANSWERT_i=0
In cyclic exponential smoothing, L represents..., - ANSWERThe length of the cycle or
season
In cyclic exponential smoothing, C_1 ... C_L = ___?, - ANSWER1. In other words,
initialize it to no initial cycle.
Exponential, trending and cyclic smoothing are also referred to as - ANSWERsingle
double and triple.
Triple smoothing is also known as? - ANSWERWinter's or Holt-Winter's
What is the optimization formula for Exponential Smoothing? - ANSWERmin(F_t-
Xt)^2 where alpha and beta are between 0 and 1.
ARIMA stands for? - ANSWERAutoregressive Integrated Moving Average
Exponential smoothing is an order ___ autoregressive model. - ANSWERInfinity. It
uses data going all the way back.
For ARIMA, the D parameter is used to specify ___. - ANSWER, The order, or the
differences of the differences of the differences (d-times.)
For ARIMA, the P parameters is used to specify ____., - ANSWERThe order of
periods (autoregression).
For ARIMA, the Q parameter is used to specify ______., - ANSWERThe order of the
moving average.
ARIMA(0,1,1) is ?, - ANSWERExponential smoothing.
, What is the order of the ARIMA parameters? - ANSWERp d q
GARCH estimates what? - ANSWERVariance.
Variance can be a proxy for ___ or ___. - ANSWERvolatility or Risk
What parameter does GARCH not have the ARIMA has? - ANSWERd because
GARCH doesn't deal with differences.
What is a simple linear regression? - ANSWERLinear regression with one predictor.
A linear regression defines a relationship between what and what? -
ANSWERPredictor(S) and Response
What is the formula for a linear regression with M predictors? -
ANSWERY=a0+sum(j=1..m, a(j)*(x(j))
What measure is used to determine the quality of a linear regression line to data? -
ANSWERSquare of the difference between the line and the data points. (Sum
squared error.)
What is the formula for point error in linear regression? - ANSWERy(i)-Yhat(i)=y(i)-
(a(0) + a(i)*x(i)
Taking partial derivates and setting them equal to zero and then solving that system
equation helps us do what? - ANSWERMinimize the error and optimize the
coefficients for linear regression.
What does AIC stand for? - ANSWERAkaike Information Criterion
What is "Likelihood"? - ANSWERA measure for the probability density for any
parameter set.
What is Maximum likelihood? - ANSWERParameters that give the highest
probability.
What is MLE? - ANSWERMaximum Likelihood Estimate... The set of parameters
that minimizes the sum of square errors.
What is the formula for AIC? - ANSWERAIC=2k - 2*ln(L*) where L* is the maximum
likelihood value and K is the number of parameters estimated.
What is the penality term in AIC and what does it do? - ANSWER2K - It helps
prevent overfitting.
A models that is fit to random effects and not real ones is said to be? -
ANSWEROverfit
