Lecture 11

Question 1

Describe two types of Use Cases for Supervised Machine Learning. Give examples of each type.

Answer 1

Recommendation: Predict which alternatives a user would prefer (Product recommendation, job recruiting, Netflix Prize, online dating, content recommendation).

Imputation: Infer the values of missing input data (Incomplete patient medical records, missing customer data, census data).

Question 2

Discuss two advantages and two challenges of Machine Learning.

Answer 2

Advantages -
Accurate: more data, the accuracy can increase automatically.
Automated: learn new patterns automatically.

Challenges -
Acquiring data in a usable form.
Formulating the problem so that machine learning can be applied, and will yield a result that’s actionable and measurable.