Studied by 2 people
Looks like no one added any tags here yet for you.
Regression
With known numerical outcomes, can we predict outcomes for data
Classification
With known groups, how can we classify new data
Dimensionality Reduction
Can I capture the patterns in my data with fewer dimensions
Clustering
Are there meaningful groups in the data
Supervised Data
Type of data where each input is paired with a corresponding output label
Unsupervised Data
Type of data where inputs do not have corresponding output labels
Model
Simplified representation of reality
Observations
Feature (x) and target (y) values
Predictions
Target (y) values calculated using regression equation
Ordinary Least Squares
Method to solve for parameters in the regression equation by using observations we have. Minimizes the sum of squared errors
