KPMG Unsupervised K Means and PCA

What is K-means clustering used for?

• a) Classification

• b) Regression

• c) Unsupervised clustering

• d) Supervised learning

In the K-means algorithm, what is the purpose of the "K" value?

• a) Number of data points

• b) Number of clusters

• c) Number of iterations

• d) Number of dimensions

Which of the following is true about the K-means algorithm?

• a) It is a supervised learning algorithm.

• b) It is used for regression tasks.

• c) It aims to minimize intra-cluster variance.

• d) It requires labeled training data.

How is the initial centroid position chosen in the K-means algorithm?

• a) Randomly from the dataset

• b) As the mean of all data points

• c) Equidistant from all data points

• d) Based on class labels

What is the objective function in the K-means algorithm?

• a) Sum of squared distances between data points and centroids

• b) Sum of absolute differences between data points and centroids

• c) Sum of cosine similarities between data points and centroids

• d) Sum of Euclidean distances between data points and centroids

Which of the following scenarios is suitable for using the K-means algorithm?

• a) Image classification

• b) Predicting stock prices

• c) Customer segmentation

• d) Spam email detection

What does the term "inertia" refer to in the context of the K-means algorithm?

• a) The rate of convergence

• b) The distance between data points and centroids

• c) The number of clusters

• d) The number of iterations

What is PCA primarily used for?

• a) Feature selection

• b) Data visualization

• c) Dimensionality reduction

• d) Outlier detection

In PCA user can put all the variance of the data into 3 or 4 components

• True

• False

• May be

• None of the above

PCA can be used for bigger datasets to reduce their dimensions and perform operations

• True

• False

• May be

• Don't know