Business Case

Question 1

What are the 3 preprocessing steps and why we do them?

Answer 1

1. Balance Dataset ->
2. Divide the dataset into training, validation, and test -> prevent overfitting
3. Save the data into a tensor friendly format

Question 2

What are the steps in successfully achieving this business case?

business case action plan

Answer 2

1. Preprocess the data
2. Create the machine learning algorithm

Question 3

What is balancing a dataset?

Answer 3

Balance the targets so the ML algorithm can learn the difference between both outcomes

Question 4

What are priors?

Answer 4

The percentage of outcomes for each class

example: if a dog / cat classification dataset has 80% cats and 20% dogs then the priors would be:
cats: 0.8
dogs: 0.2

Question 5

When is a dataset balanced?

Answer 5

When you have a 50% / 50% of the priors