• Introduce ourselves
• Introduce the course
• Go through the syllabus
• Talk about some introductory concepts

• Please mute your sound while in the main meeting

• In breakout rooms, unmute and turn your video on if possible

• If you have a question or comment, type in the chat

• You can also raise your hand virtually in the participants tab

Please be aware that the class is being recorded

• Name
• Major
• Hometown
• Interest in SDS 323
• …

• A “new” course offered by SDS since 2018
• Intersection of statistical inference and machine learning
• Aim to teach you how to use methods/tools, not how to invent or build one
• Understand meaning, strengths and weaknesses of various models, not the underlying theories
• Hands-on experience via the statistical software \(\texttt{R}\)

Materials will be posted on Canvas:

• Syllabus
• Lecture Slides
• Homework
• R code

If you are on WL, talk to me at the office hours after class!

• Homework assignments can be done in groups with up to 4 people
• If you already know some of your peers and want to create a group, e-mail me before next Tuesday
• If you do not know anyone, do not worry! I will manually create groups of 4 people and will do my best to make them heterogeneous so that you will have different strengths

• Motivation to participate in class
• Excitement about homework
• Time management
• Interest in your own set of problems

What .gif best describes your feelings towards this semester?

A set of scientific tools to collect data, and turn data into information, knowledge, evidence and insights. It should result in better understandings, judgments and decisions about a phenomenon

What is that we “learn”, in Statistics?

• Patterns
• Correlations
• Make predictions
• …

Predicting, or estimating, an output \(y\) based on one or more inputs \(x\).

• Given past data on outcomes \(y\) paired with features \(x\), can we find patterns that allow us to predict \(y\) using \(x\)?
• Key characteristic: there is a single privileged outcome \(y\)
• Example: a house has \(3\) bedrooms (\(x_{1}\)), \(2\) bathrooms (\(x_{2}\)), \(2100\) square feet \((x_{3})\), and is located in Hyde Park (\(x_{4}\)). What price (\(y\)) should it sell for?

• We still have multivariate data (many input variables \(x\)) and want to find patterns.
• But there is no single privileged outcome
• Example: "Here’s data on the shopping basket of every Whole Foods customer at \(6^{th}\) and Lamar last month. Find some patterns that we can use to improve product placement’’

The first two weeks might feel slow paced if you have solid prerequisites. I want to give you time to properly learn the foundations that we will need in the rest of the course.

• Tuesday 09/01: introduction to probability
• Thursday 09/03: first hands on experience with R
• Tuesday 09/08: introductory topics

The first homework will cover the prerequisites and make sure that we can start building on them. Start working on it early!