Question-21: Why do you say, unsupervised Machine Learning is data driven?

Answer: Because unsupervised learning is based upon the data and characteristics/features/properties of the data. And also, the outcome from an unsupervised learning tasks are controlled by these data. Hence, it is called data driven Machine Learning.

Question-22: Is recommender system is part of unsupervised learning or supervised learning?

Answer: This would fall into the un-supervised learning. Let’s take an example of online video platform like YouTube. On YouTube following characteristics are known in advance.

• Title of Video.
• Length of video.
• Genre of the video.

And history of the videos watched by a user. Taking all these data into account, YouTube needs to recommend or suggest new videos that you might have not seen yet. So recommender system will try to find the relationship in the data and based on that it will suggests you new video to watch.

Question-23: Can you give couple of more example for un-supervised learning?

Answer: Below are the examples of unsupervised learning

• Grouping customer based on buying habits.
• Grouping user complains.

Question-24: In one line can you tell me the difference between supervised and un-supervised learning?

Answer: Yes, Supervised learning has the label already known to predict. But Un-supervised does not know in advance. And these labels are also known as dependent variables.

Question-25: Email Spam detection is what kind of classifications?

Answer: You, can consider email spam detection is as a Binary Classification under supervised learning. Because, here you are trying to find that particular email is a spam or not. Only one of the two possible labels.