Unsupervised Machine Learning Concepts

The video introduces unsupervised machine learning, focusing on clustering techniques like k-means and hierarchical clustering. It explains how these methods can group data without predefined labels, using examples like pizza ordering habits and Autism Spectrum Disorder. The video also discusses evaluating clusters with silhouette scores and highlights the practical applications of clustering in various fields.

What is the main goal of unsupervised machine learning?

• To create labels for existing data

• To improve the accuracy of supervised models

• To predict future data points

• To classify data into existing categories

In the k-means clustering example, what is the purpose of centroids?

• To represent the center of each cluster

• To act as the initial data points

• To determine the number of clusters

• To serve as the final output of the algorithm

How does the k-means algorithm determine when to stop iterating?

• When the silhouette score is maximized

• When all data points are assigned to a cluster

• When the centroids and groups stop changing

• When the number of clusters reaches a predefined limit

What does a high silhouette score indicate about a cluster?

• The data points within the cluster are close to each other

• The cluster is not well-defined

• The cluster is very similar to other clusters

• The cluster contains a large number of data points

What is a dendrogram used for in hierarchical clustering?

• To assign data points to clusters

• To determine the number of clusters

• To calculate the silhouette score

• To visualize the distance between clusters

In hierarchical clustering, what does it mean when two clusters join higher up in the dendrogram?

• They have higher silhouette scores

• They are less similar to each other

• They contain more data points

• They are more similar to each other

How can hierarchical clustering be beneficial in Autism Spectrum Disorder research?

• By predicting the severity of the disorder

• By eliminating the need for traditional therapies

• By creating profiles for targeted therapies

• By diagnosing new cases of ASD

What is the main difference between k-means and hierarchical clustering?

• K-means is more accurate than hierarchical clustering

• K-means is used for supervised learning, while hierarchical is for unsupervised learning

• K-means creates flat clusters, while hierarchical reveals subgroups

• K-means requires a predefined number of clusters, hierarchical does not

What is the purpose of using radar graphs in the ASD research example?

• To visualize the number of clusters

• To display the developmental domain scores

• To calculate the silhouette score

• To determine the severity of ASD

What type of clustering starts with each data point as its own cluster?

• Density-based clustering

• Agglomerative clustering

• Divisive clustering

• K-means clustering