1What is clustering?
An unsupervised learning algorithm that groups similar data points into clusters
2What does clustering aim to find?
Similarity and relationship patterns among data points
3Is clustering a predictive technique?
No, but it is valuable in the predictive analytics process
4What is the main objective of clustering?
To identify underlying patterns and relationships within a dataset
5What are examples of features that clustering can use?
Shape, size, color, behavior, usage, etc.
6What is an example of clustering in supermarkets?
Grouping similar items like groceries, hardware, and clothing into distinct departments
7What are some types of clustering?
Partitioning, density-based, hierarchical, model-based, fuzzy clustering
8What algorithm is used in partitioning clustering?
K-means algorithm
9What does density-based clustering do?
Divides areas of high-density data points into clusters
10What is hierarchical clustering?
An alternate to partitioned clustering that creates a tree-like structure called a dendrogram
11What are some common applications of clustering?
Customer segmentation, social network analysis, image segmentation, anomaly detection
12What algorithm is commonly used for hierarchical clustering?
Agglomerative hierarchical algorithm
13What are popular clustering algorithms in machine learning?
K-means, Mean-shift, DBSCAN, Expectation-Maximization with GMM
14How is clustering used in text analytics?
It groups documents by topics or authors
15What is a decision tree algorithm?
A supervised learning method that splits the dataset on its predictor variables based on different conditions.
16What tasks are decision trees used for?
They are used for both classification (discrete) and regression (continuous) tasks.
17What are the two main entities of a decision tree?
Decision nodes (where the data is split) and leaves (the final outcomes).
18What is a classification decision tree?
A type of decision tree that uses discrete predictor variables to answer simple discrete questions with categorical or discrete outcomes.
19What is a regression decision tree?
A type of decision tree that uses continuous predictor variables and produces numerical or continuous outcomes.
20What is the Gini method used for in decision trees?
It is used to generate binary splits on the attributes (predictor variables) of the dataset.
21What is entropy in decision trees?
A measure of disorder or uncertainty in the data, used to calculate how much information is gained from a question.
22What is information gain?
The reduction in uncertainty when a question is asked, helping predict the target variable better.
23What is Shannon Entropy?
A measure that represents the amount of uncertainty or randomness in data, denoted as H(S).
24What happens when the entropy is 1.0?
There is high uncertainty, meaning the outcome is unpredictable, like when tossing a fair coin.
25What happens when the entropy is 0?
There is high certainty, meaning the outcome is completely predictable, like tossing a coin with heads on both sides.
26What is the first step in creating a decision tree?
Create the root node for the tree.
27When does a decision tree return a positive leaf node?
When all cases are positive.
28What is the stopping condition for a decision tree?
When the tree has all leaf nodes or runs out of attributes.
29What is a random forest?
A technique to avoid overfitting by generating multiple decision trees and averaging their outputs.
30What is an example of decision tree use?
Making decisions about whether to approve or reject a loan application.