EE2211 Intro to ML
Tutorial #1 (T03)
Juezhao Yu (Jay)
[email protected]
©©Copyright
CopyrightNational
NationalUniversity
UniversityofofSingapore.
Singapore.All
AllRights
RightsReserved.
Reserved.
Logistics
• Mon 14:00 ~ 16:00, EA-06-06 (T03)
• 12 tutorials in total
• Tutorial attendance accounts for 4% of final score
• Absence permission granted on a case-by-case manner (email as a
formal record)
• On Week 2 and Week 3, we run trials on the attendance marking
system. (Everyone marked as attended)
• On Week 4, we start logging your attendance.
© Copyright National University of Singapore. All Rights Reserved.
13th Nov (PH):
A recorded Zoom session on Week 13 to make up.
Timetable Attendance is optional.
Will let you know more on Week 12.
Tutorial-Q1
… Release of Asgn1
Tutorial-Q5 Release of Asgn2
Mid-term, 26th Sep
Tutorial-Q6
Release of Asgn3
…
Tutorial-Q12
© Copyright National University of Singapore. All Rights Reserved. Final, 27th Nov
Contact
• Email: [email protected]
• Title: [EE2211] XXXXXXX
• Office hours: after each tutorial session
© Copyright National University of Singapore. All Rights Reserved.
Dos
• For math-related questions, do approach me if you
• Have doubts on lecture notes;
• Have difficulties in solving math questions in tutorials.
• For python-related questions, do approach me if you
• Get stuck on a coding problem for over 2 days.
• For choice of topics, do indicate your interests via email.
• If that’s a majority choice, I will include in our discussion in tutorial
sessions.
• If conditions are not met (e.g., few emails received, limited time in
class), I will make a side note for you to study by yourselves. Also, it’s
possible to make Zoom meetings to take questions.
© Copyright National University of Singapore. All Rights Reserved.
Communication
• Communicating to yourself
• Ask yourself the right and important questions
• (which is probably THE most important step for research)
• Communicating with me (the tutor)
• Ask questions efficiently!
• I appreciate questions of good quality.
• Communicating with your peers
• You are encouraged to form small study group!
• Communicating with AIs (prompt engineering)
© Copyright National University of Singapore. All Rights Reserved.
Warming Up
• Raise your hands if you have done the following things,
• Written a computer program in Matlab/VB/Python/…?
• (“Hello World!” counts!!!)
• Visualized data with Matlab/Excel/plt.plot/…?
• Managed data with Excel/Pandas/…?
• Done matrix calculation in Matlab/Numpy/…?
© Copyright National University of Singapore. All Rights Reserved.
Tutorial 1 – Questions 1, 2, 3, 8
• Definition of Machine Learning and Artificial Intelligence.
• AI: let computers think like humans
• ML: let computers acquire human-defined knowledge
• Task T
• Experience E
• Performance measure P
© Copyright National University of Singapore. All Rights Reserved.
Patient ID Fever? Cough? Tired? Sore Throat? … X-Ray Scan COVID-19?
#1 0 0 1 1 … ./xray-image/1.jpg 1
#2 1 1 0 0 … ./xray-image/2.jpg 0
#1000 1 1 1 1 ./xray-image/1000.jpg 1
Input Output
© Copyright National University of Singapore. All Rights Reserved.
Tutorial 1 – Questions 4, 5, 6
• Classification vs. Regression
• Supervised Learning
• For each data sample, we need to assign a value to it
• Labels (discrete values): classification
• Continuous values: regression
© Copyright National University of Singapore. All Rights Reserved.
© Copyright National University of Singapore. All Rights Reserved.
Tutorial 1 – Questions 7, 9
• Supervised learning vs unsupervised learning
• If there is a 𝑦 (ground truth) for each 𝑋 (sample) : supervised learning
𝑓 𝑋 = 𝑦% ⟷ 𝑦 ⇒ min 𝑓 𝑋 − 𝑦
!
• If ground truth is absence: unsupervised learning (usually to reveal
patterns among samples)
© Copyright National University of Singapore. All Rights Reserved.
Clustering
© Copyright National University of Singapore. All Rights Reserved.
Prior knowledge: different types of heart disease, etc.
© Copyright National University of Singapore. All Rights Reserved.
Tutorial 1 – Question 10
• Induction and deduction
• Induction: mass observation -> pattern recognition
• Deduction: a chain of logic
Inductive
Deductive
© Copyright National University of Singapore. All Rights Reserved.
For Next Week …
• We will do some coding together.
• Please bring along your laptops with a ready Python
environment.
• We will talk about:
• Data processing with Pandas
• Data visualization with matplotlib.pyplot
© Copyright National University of Singapore. All Rights Reserved.
See you next week!
© Copyright National University of Singapore. All Rights Reserved.
