Unsupervised Learning

Question 1

What is an unsupervised learning algorithm?

Answer 1

An algorithm that doesn’t require the true labels of the data.

Question 2

What are some real uses of unsupervised learning?

Answer 2

1. Clustering
2. Topic Modelling
3. Community Detection

Question 3

What are the main steps to clustering?

Answer 3

1. Go over all datapoints
2. Measure the distance or similarity between points
3. Identify clusters of points that are closer to each other than they are to points outside the cluster.

Question 4

What are two examples of clustering algorithms?

Answer 4

1. K-means
2. DBSCAN

Question 5

How does K-mean clustering work?

Answer 5

It identifies points close to k centroids, where K is given by the user

Question 6

How does DBSCAN work?

Answer 6

It finds core regions of high density and expands clusters form them

Question 7

What are two types of clustering?

Answer 7

Hard Clustering and Soft Clustering

Question 8

What is hard clustering?

Answer 8

Each object belongs in one cluster, similar to how a perceptron performs classification

Question 9

What is soft clustering?

Answer 9

Objects are assigned to multiple clusters with corresponding probabilities.

Question 10

What is Hierarchical Clustering?

Answer 10

A clustering algorithm that assigns a data point to multiple clusters, assigning them to all the clusters of their parent nodes.