The goal of supervised machine learning is to: a. Find patterns in data* b. Continuously improve a model c. Create predictions about future events d. Generate insights about a dataset

A deep learning architecture consists of: a. Feature analysis and data preprocessing b. Neural networks that are trained sequentially c. An input layer, one or more hidden layers, and an output layer * d. An optimization algorithm and a loss function

What are the benefits of transfer learning? a. It requires minimal data cleansing b. It allows you to use existing architectures * c. It is faster than training a model from scratch d. It is more accurate than training a model from scratch

When performing logistic regression, what is the goal of the logistic function? a. To create a linear line that best fits the data b. To transform the output of a linear equation into values between 0 and 1 * c. To calculate the optimal parameters for classification d. To classify observations into one of two groups

In supervised machine learning, what is used to measure the accuracy of a model? a. The number of training examples b. The number of data preprocessing steps c. The classification rate or the mean squared error * d. The amount of computing power used

What is an example of a supervised learning task? a. Clustering b. Image classification * c. Text summarization d. Detecting anomalies

What type of learning is required to build a recommendation system? a. Semi-supervised b. Reinforcement * c. Unsupervised d. Transfer

What is k-nearest neighbors (KNN) used for? a. Identifying features of a dataset b. Reducing data dimensions c. Classifying examples based on the samples that are closest to them * d. Generating insights about a dataset

What is the purpose of regularization? a. To reduce the complexity of a model * b. To improve the accuracy of a model c. To reduce data dimensionality d. To identify patterns in data

What is cross-validation used for? a. To detect overfitting b. To ensure all data points are used in training * c. To reduce the variance of a model d. To compare performance between models