CSE 163

Machine Learning�

Suh Young Choi�

🎶 Listening to: Mariusz Duda

💬 Before Class: Do you have a favorite tree?

​

Last Time

• Statistics 101
• Hypothesis testing
• Best practices for research

This Time

• Machine Learning
• Terminology
• Types of ML
• ML Code (scikit-learn)
• Decision Trees

2

Updates & Reminders

Reading Assignment 4 is out on Canvas/Hypothesis now

• Due Tuesday, 2/17

​

THA 4 is due tomorrow (2/12)

• If you need to prioritize anything, focus on the Creative Component!

​

Checkpoints and Reading Assignments will open for submissions after initial grades are released

​

​

Suh Young will not be here next week!

• TAs are hosting a watch party of pre-recorded lectures instead

​

​

3

Checking in

Advanced material / Out-of-scope deductions

• The expectation is that submitted work is aligned with course material and guidelines
• The more we see it in your submission, the larger the deduction becomes
• Unresolved cases result in a 0 + forfeiture of resubmission ☹

​

​

Resubmission Policy

• Two resub cycles are open after the initial round of feedback per assessment
• You can resubmit one technical component per cycle!
• Exact dates for final submission days will be posted to Ed

4

Terms

• Machine learning
• Model
• Machine learning algorithm
• Data
• Training Set
• Example
• Feature
• Label
• Types of ML
• Regression
• Classification

5

Features Matter

• There exists hundreds of different model types, but the most important thing is having good features to describe your data
• Garbage In => Garbage Out

​

6

Decision Tree

• Like a series of GIANT if/else branches! (Hence… tree)

7

Code Recap�(classification)

General ML pipeline (at least to start)

• For classification tasks

8

# Separate data

features = data.loc[:, data.columns != 'target']

labels = data['target']

# Create and train model

model = DecisionTreeClassifier()

model.fit(features, labels)

# Predict on some data

predictions = model.predict(features)

# Assess accuracy

accuracy_score(labels, predictions)

​

Code Recap�(regression)

General ML pipeline (at least to start)

• For regression tasks

9

# Separate data

features = data.loc[:, data.columns != 'target']

labels = data['target']

# Create and train model

model = DecisionTreeRegressor()

model.fit(features, labels)

# Predict on some data

predictions = model.predict(features)

# Assess MSE

mean_squared_error(labels, predictions)

​

Group Work:

Best Practices

When you first working with this group:

• Introduce yourself!
• If possible, angle one of your screens so that everyone can discuss together

​

Tips:

• Starts with making sure everyone agrees to work on the same problem
• Make sure everyone gets a chance to contribute!
• Ask if everyone agrees and periodically ask each other questions!
• Call TAs over for help if you need any!

10

Next Time

• Evaluating Machine Learning

​

Before Next Time

• Complete Lesson 15
• Remember not for points, but do go towards Weekly Tokens
• THA 4 due tomorrow night!
• Get started on EDA or Milestone!

11