FA2 - Machine Learning

The larger variety of data points your data set contains, the more complex a model you can use without overfitting. (T or F)

True

Binary Classification is a classification of dichotomous classes. (T or F)

True

The _____ allows the models to make informed predictions even when faced with previously unseen data. 

Underfitting

Generalization

Overfitting

Generalization

Supervised algorithms address classification problems where the output variable is categorical. (T or F)

False

The ______ refers to the error resulting from sensitivity to the noise in the training data.

variance

The more complex we allow our model to be, the better we will be able to predict on the training data. (T or F)

True

SVM is an example of regression algorithm. (T or F)

False

In k-NN, High Model Complexity is overfitting. (T or F)

True

In k-NN, when you choose a small value of k (e.g., k=1), the model becomes less complex. (T or F)

False

The ‘k’ in k-Nearest neighbors refers to an arbitrary number of neighbors. (T or F)

True

In k-NN, voting means for each test point, we count how many neighbors belong to a class e.g. how many belong to class 0 and how many neighbors belong to class 1.  (T or F)

True

In the estimation of regression model, predicting worse than the average can result in negative numbers. (T or F)

True

When comparing training set and test set scores, we find that we predict very accurately on the training set, but the R2 on the test set is much worse. This is a sign of overfitting. (T or F)

True

Lasso uses L2 Regularization. (T or F)

False

What is the full form of OLS?

Ordinary Least Squares

Regularization means explicitly restricting a model to avoid overfitting. (T or F)

True

Ridge is generally preferred over Lasso, but if you want a model that is easy to analyze and understand then use Ridge. (T or F)

False

In Ridge regression is α (alpha) is larger, the penalty becomes larger. (T or F)

True

Naive Bayes classifier assumes that the presence of a particular feature in a class is unrelated to the presence of any other feature. (T or F)

True

Naïve Bayes classifier that deals with continuous data.

All the given options

GaussianNB

MultinomialNB

BernoulliNB

GaussianNB

Its target is a categorical variable.

Correlation

Supervised Learning

Regression

Classification

Classification

Regression predicts consecutive numbers. (T or F)

True

In k-NN, Low Model Complexity is underfitting. (T or F)

True

In k-NN, Low Model Complexity is overfitting. (T or F)

False

The ‘offset’ parameter is also called intercept. (T or F)

True

Types of Linear Models : Linear Regression, ____________. 

Logistic Regression

The ‘slope’ parameter is also called weights or coefficients. (T or F)

True

Naïve Bayes classifier that deals with integer count data.

MultinomialNB

All the given options

BernoulliNB

GaussianNB

MultinomialNB

A model which does not capture the underlying relationship in the dataset on which it's trained.

Generalization

Overfitting

Underfitting

Underfitting

A model is able to make accurate predictions on new, unseen data.

Underfitting

Overfitting

Generalization

Generalization

When using multiple nearest neighbors, the prediction is the mean of the relevant neighbors. (T or F)

True

The ‘offset’ parameter is also called _______.

Intercept

Slope

Weights

Mean

Intercept

In Ridge regression is α (alpha) is larger, the penalty becomes lesser. (T or F)

False

Naïve Bayes classifier that deals with integer binary data.

BernoulliNB

GaussianNB

d. All the given options

MultinomialNB

BernoulliNB

Naïve Bayes learns parameters by looking at each feature individually and collects simple per-class statistics from each feature. (T or F)

True