Introduction to Machine Learning Course at University of Washington, Lecture notes of Machine Learning

Download Introduction to Machine Learning Course at University of Washington and more Lecture notes Machine Learning in PDF only on Docsity!

CSE/STAT 416

Introduction + Regression

Vinitra Swamy University of Washington June 22, 2020 Slides and materials for this course courtesy of Hunter Schafer and Emily Fox.

Machine Learning is

changing the world.

It’s Everywhere! Retail Movie Distribution Music Advertising Networking Search Dating Legal Advice Human Resources Productivity Software Campaigning (^) Real Estate Wearables

CRM

Disruptive companies differentiated by INTELLIGENT APPLICATIONS using Machine Learning

It’s Everywhere…

What is Machine Learning? Generically (and vaguely) Machine Learning is the study of algorithms that improve their performance at some task with experience

Data

ML

Method

Intelligence

Course Topics

• Linear regression, regularized approaches (ridge, LASSO)
• Linear classifiers: logistic regression
• Non-linear models: decision trees
• Nearest neighbors, clustering
• Recommender systems
• Deep learning Models
• Gradient descent
• Boosting
• K-means Algorithms
• Point estimation, MLE
• Loss functions, bias-variance tradeoff, cross-validation
• Sparsity, overfitting, model selection
• Decision boundaries Concepts

ML Course Landscape

CSE 446

▪ CSE majors ▪ Very technical course STAT 435 ▪ STAT majors ▪ Very technical course CSE/STAT 416 ▪ Everyone else!

• This is a super broad audience! ▪ Give everyone a strong foundational understanding of ML
• More breadth than other courses, a little less depth

STAT

CSE

CSE/STAT

Course Logistics

Who am I? ▪^ Vinitra Swamy

• Lecturer, Paul G. Allen School for Computer Science & Engineering (CSE)
• AI Software Engineer at Microsoft
  • AI Framework Interoperability
  • Open Neural Network eXchange (ONNX) ▪ Office Hours
• Time: 4:00 pm - 5:00 pm, Fridays, or by appointment
• Location: Zoom ▪ Contact
• Personal Matters: vinitra@cs.washington.edu
• Course Content + Logistics: Piazza

16 Lecture Mon Nothing Tue Lecture Wed Section Thur Nothing Fri ● We happen to not record attendance in lectures and section, but attending these sessions is expected ● Participation component (5% of your grade) ● Zoom Recordings for Lecture (on Canvas) Previous HW Due Next HW Released

Lectures Introduced to material for the first time. Mixed with activities and demos to give you a chance to learn by doing. No where near mastery yet! Sections Practice material covered in 1 in a context where a TA can help you. The emphasis is still on you l earning by doing.^2 Homework With the scaffolding from 1 and 2 , you are probably now capable to tackle the homework. These will be complex and challenging, but you’ll continue to learn by doing. 1 3

Homework Logistics ● Late Days ○ 4 Free Late Days for the whole quarter. ○ Can use up to 2 Late Days on any assignment. ○ Each Late Day used after the 4 Free Late Days results in a - 10% on that assignment ● Collaboration ○ You are encouraged to discuss assignments and concepts at a high level ■ If you have code in front of you in your discussion, probably NOT high level ■ Discuss process, not answers ○ All code and answers submitted must be your own ● Turn In ○ Everything completed and turned in on Gradescope ○ Multiple “assignments” on Gradescope per assignment

Case Study 1 Regression: Housing Prices