Unsupervised learning Flashcards
Q: What is the second most widely used form of machine learning after supervised learning?
A: Unsupervised learning.
Q: How does unsupervised learning differ from supervised learning?
A: In unsupervised learning, the data is not associated with any output labels (Y). The algorithm must find structure or patterns in the data on its own.
Q: What is a clustering algorithm in the context of unsupervised learning?
A: A type of unsupervised learning algorithm that groups data into different clusters based on patterns or similarities within the data.
Q: Give an example of how clustering algorithms are used in Google News.
A: Clustering algorithms group related news articles that mention similar words (like “panda,” “twin,” and “zoo”) into clusters of related stories.
Q: How does Google News use clustering algorithms without supervision?
A: The algorithm figures out on its own which words suggest that certain articles are in the same group, without any human intervention.
Q: What is another example of unsupervised learning applied to clustering genetic or DNA data?
A: By analyzing DNA microarray data, the algorithm can group individuals into different types based on genetic expression, such as Type 1, Type 2, and Type 3 individuals.
Q: Why is clustering genetic data considered unsupervised learning?
A: Because the algorithm is not given predefined labels or types, it must discover the structure and categorize individuals based on the data alone.
Q: How can clustering be applied to customer data in business?
A: Companies can use clustering to automatically group customers into different market segments to better understand and serve them efficiently.
Q: What is an example of market segmentation using clustering algorithms?
A: The DeepLearning.AI team used clustering algorithms to identify different groups in their community, such as those seeking knowledge, career advancement, or staying updated in their field.
Q: What does a clustering algorithm do with data without labels?
A: It tries to automatically group the data into clusters based on inherent patterns or similarities.
Q: Are there other types of unsupervised learning besides clustering?
A: Yes, there are other types of unsupervised learning algorithms beyond clustering.
Q: Why is unsupervised learning called “unsupervised”?
A: Because it does not involve providing the algorithm with labeled outputs to learn from; the algorithm must find patterns or structures in the data on its own.
Q: What does an unsupervised learning algorithm need to do without output labels?
A: It has to find some structure, pattern, or something interesting in the data.
Q: What is one type of unsupervised learning used for fraud detection?
A: Anomaly detection, which identifies unusual events or transactions.
Q: What is dimensionality reduction in unsupervised learning?
A: A technique that compresses a large dataset into a much smaller one while retaining as much information as possible.
