City College, Fall 2018

Intro to Data Science

Week 5: Intro to Linear Models

Today's Agenda

1. Linear Regression
2. Assumptions for Linear Models
3. Measuring Performance for Linear Models
4. Regression vs. Classification
5. Logistic Regression

Week 4 Recap

1. Types of Data
2. Useful Statistical Distribution
3. Important Summary Statistics
4. Key Theorems

HW Recap

1. How was DataCamp?
2. Project Updates
• What are the key relationships from the data sets you've selected?
• How might you apply a model to your data?

Data Science Models

You try.

Assumptions of Linear Regression

1. Data is linear in form.
2. Sample is random.
3. Error terms have constant variance (homoscedasticity).
4. Error terms have a mean of zero based on the observed data.
5. Predictors are independent (no multicollinearity).
6. Errors are normally distributed.

Data is linear in form.

Sample is random.

Error terms have constant variance (homoskedasticity).

Errors are uncorrelated.

Predictors are independent (no multicollinearity).

in the case of:

Errors are normally distributed.

Output

Key Metrics of Success

1. R-squared
2. Adjusted R-squared
3. Coefficients
4. P-values

R-Squared

Adjusted R-squared

Coefficients

P-values

Does linear regression prove causality?

This Week's Data

Assignment 5: Due Monday, October 15 by 6:30pm

Part I: DataCamp

• This week's DataCamp assignment covers two chapters of Supervised Learning with scikit-learn, but does not require the full course. It should be shorter than past DataCamp assignments and will appear in your DataCamp account.

• The assignment notebook has a few short questions building on today's lecture and data exercise. Please complete the exercises as instructed in each individual notebook cells to obtain full credit for the assignment. To submit your assignment, please submit a completed jupyter notebook (*.ipynb files only) through Blackboard.

Project Proposal: Due Friday, October 19 by 11:59pm

Details on the project are now available here.