Lecture 22
For Today You Should Have:
1) Filled out code to make a learning curve in the iPython notebook from last time
Today:
1) More ML
For next time
1) Prepare for a quiz
More ML
Overview of last time:
ML can be thought of as programming by examples: an ML algorithm converts a list of desired input / output pairs into a program (or a model) that can predict the output for a novel input.
The data we use to create the model is called the training set. To evaluate performance on novel input, we typically use an additional set of data called the testing set to measure expected performance.
Many of the techniques you have learned can be thought of as machine learning algorithms (e.g. linear / logistic regression).
Today we are going to get down into the details of how to apply machine learning.