Machine Learning Exam 1 CS 4375
Verified 2024
Active algorithm - ANSWER-Reinforcement Learning teaches active agents to identify
optimal actions given the current environment and what has been learned in past
experience
additive assumption - ANSWER-each predictor contributes to the model independently
of other predictors.
AIC - ANSWER-Akaike Information Criterion is useful for comparing models. It
penalizes overly complex models.
Array - ANSWER-A n-dimensional object with the same type.
AUC - ANSWER-Area under the curve for a ROC curve quantifies the predictive value.
A value of .5 means no predictive value, 1 being a perfect classifier.
batch learning - ANSWER-train algorithm on a portions of data, then using the
remaining data to test and evaluate how well the trained model can perform on
previously unseen data. Named batch because data is fed in one batch.
Classification - ANSWER-the target is qualitative, a class, like predicting if a borrower is
a good credit risk or not, given their income, outstanding credit balance, and other
predictors.
column, attribute, feature, predictor - ANSWER-
Confounding variable (interaction effect) - ANSWER-a variable that correlate with both
the target and a predictor.
Dataframe - ANSWER-A 2d structure where each column may be a different type.
Deviance residual - ANSWER-a mathematical transformation of the loss function, and
quantifies a given point's contribution to the overall likelyhood. (similar to a RSS stat in
Linreg)
Difference between linear and logistic regression - ANSWER-Whereas the coefficient of
a linear regression predictor quantifies the difference in the target variable as the
predictor changes, in logistic regression, the coefficient quantifies the difference in the
log odds of the target variable.
, F-statistic - ANSWER-Takes into account all of the predictors to determine if they are
significant predictors of Y. It proves evidence against the null hypothesis that the
predictors are not really predictors.
The advantage of the F-statistic over R^2, is that R^2 does no tell us whether it is
statistically significant or not but the F-statistic does.
F-statistic greater than 1, and a low p-value indicate confidence in our model.
Fields that make up machine learning - ANSWER-AI, Stat, CS, Probability
Gradient Decent - ANSWER-One of the most commonly used optimization techniques.
Will not be bogged down for any large datasets.
The algorithm starts with some value for the parameter w and keeps changing them in
an interactive loop until it finds a minimum.
If step size is too small it will take too long to calculate, too large and we might step over
minimum.
High bias, low variance model - ANSWER-Is likely to underfit and not capture the shape
of the data. Happens with simpler models such as linear and logistic regression.
Informative algorithm - ANSWER-They provide information in the form of data analysis
or prediction. They provide information in the form of data analysis or prediction
Kappa - ANSWER-is a statistic that attempts to adjust accuracy by accounting for the
possibility of a correct prediction by chance alone. Closer to 1 the better agreement.
Leverage Point - ANSWER-a data point with an unusual x value
likelihood - ANSWER-It quantifies how likely it is that we would see the data given the
Survived instances
Linear Regression Strengths - ANSWER-Relatively simple algorithm with an intuitive
explanation because the coefficients quantify the effect of predictors on the target
variable.
Works well when the data follows a linear pattern.
Has low variance.
Linear Regression Weaknesses - ANSWER-High bias because it assumes a linear
shape to the data.
Verified 2024
Active algorithm - ANSWER-Reinforcement Learning teaches active agents to identify
optimal actions given the current environment and what has been learned in past
experience
additive assumption - ANSWER-each predictor contributes to the model independently
of other predictors.
AIC - ANSWER-Akaike Information Criterion is useful for comparing models. It
penalizes overly complex models.
Array - ANSWER-A n-dimensional object with the same type.
AUC - ANSWER-Area under the curve for a ROC curve quantifies the predictive value.
A value of .5 means no predictive value, 1 being a perfect classifier.
batch learning - ANSWER-train algorithm on a portions of data, then using the
remaining data to test and evaluate how well the trained model can perform on
previously unseen data. Named batch because data is fed in one batch.
Classification - ANSWER-the target is qualitative, a class, like predicting if a borrower is
a good credit risk or not, given their income, outstanding credit balance, and other
predictors.
column, attribute, feature, predictor - ANSWER-
Confounding variable (interaction effect) - ANSWER-a variable that correlate with both
the target and a predictor.
Dataframe - ANSWER-A 2d structure where each column may be a different type.
Deviance residual - ANSWER-a mathematical transformation of the loss function, and
quantifies a given point's contribution to the overall likelyhood. (similar to a RSS stat in
Linreg)
Difference between linear and logistic regression - ANSWER-Whereas the coefficient of
a linear regression predictor quantifies the difference in the target variable as the
predictor changes, in logistic regression, the coefficient quantifies the difference in the
log odds of the target variable.
, F-statistic - ANSWER-Takes into account all of the predictors to determine if they are
significant predictors of Y. It proves evidence against the null hypothesis that the
predictors are not really predictors.
The advantage of the F-statistic over R^2, is that R^2 does no tell us whether it is
statistically significant or not but the F-statistic does.
F-statistic greater than 1, and a low p-value indicate confidence in our model.
Fields that make up machine learning - ANSWER-AI, Stat, CS, Probability
Gradient Decent - ANSWER-One of the most commonly used optimization techniques.
Will not be bogged down for any large datasets.
The algorithm starts with some value for the parameter w and keeps changing them in
an interactive loop until it finds a minimum.
If step size is too small it will take too long to calculate, too large and we might step over
minimum.
High bias, low variance model - ANSWER-Is likely to underfit and not capture the shape
of the data. Happens with simpler models such as linear and logistic regression.
Informative algorithm - ANSWER-They provide information in the form of data analysis
or prediction. They provide information in the form of data analysis or prediction
Kappa - ANSWER-is a statistic that attempts to adjust accuracy by accounting for the
possibility of a correct prediction by chance alone. Closer to 1 the better agreement.
Leverage Point - ANSWER-a data point with an unusual x value
likelihood - ANSWER-It quantifies how likely it is that we would see the data given the
Survived instances
Linear Regression Strengths - ANSWER-Relatively simple algorithm with an intuitive
explanation because the coefficients quantify the effect of predictors on the target
variable.
Works well when the data follows a linear pattern.
Has low variance.
Linear Regression Weaknesses - ANSWER-High bias because it assumes a linear
shape to the data.
