Fall 23 Review

1. What is not supervised learning?

• 1. Training a model with labeled data -- giving it the expected answers!

• Grouping people into friend groups based on data about their favorite movies!

• Predicting the price of a house given historical data about cost of houses with known square footage.

• 1. Classifying cats and dogs, given pictures of each with labels.

Which of the following is true? [2 correct]

• 1. Classification problems have a continuous output line

• Classification predicts classes or categories

• An example of regression is seeing if a picture is a cat or dog

• Regression involves predicting the continuous relationship between input features and the output.

1. What kind of training data should you provide a supervised learning model for best results?

• 1. Give the model one specific type of example data only.

• 1. Give the model a lot of diverse data.

• 1. Give the model unlabeled data

• 1. Give the model one or two examples only.

What is linear regression?

• 1. The process of shifting weights and biases to minimize loss

• Moving to the left along a line to optimize the learning rate.

• Approximating a continuous (linear) output; finding the relationship between X and y

• "Finding the line of best fit"

What is loss?

• A number indicating how bad the model’s prediction was on a single example.

• 1. A function that transforms any number into a probability

• A line that shows the relationship between input features and output

• 1. A line that separates classes

Why is gradient descent important?

• 1. We want to make our predictions smaller, so descending helps

• We use it to change the weights / biases to make our model more accurate

• 1. Gradient descent increases loss. More loss helps our model get close to 0

• We use it to prevent the model from overfitting the data.

What is mean squared error? 2 correct answers!

• The absolute difference between actual value and predicted value

• 
• A loss function

• The gradient of a function

1. What do we call the line that separates two classes in logistic regression?

• Decision boundary

• Loss function

• Line of best fit