See the latest book content here.

AMSI Summer School Course 2021

These notes are used for a course taught at the Australian Mathematical Sciences Institute (AMSI) summer school in January 2021 (January = summer in Australia). This page contains basic information about the course including, schedule, communication, and assessment. Zoom links and student specific information are not open to the general public, but are rather accessible via the course’s Canvas page.

See some student project videos are below.

Schedule

The schedule specifies the times for lectures, split practicals, (optional) consultation meetings, and assesment due dates. All times are given in AEDT (NSW, Victoria, Tasmania). See further notes about the schedule below.

This table details the schedule and contains links to material and recordings that will be supplied during the course. For example, for the first split practical in hour 3, you can find links to the R practical (BL), Python practical (SM), and Julia practical (YN). There is also an additional link to original R source code.

You can also watch all YouTube recordings directly via the YouTube playlist.

Hour Week Day/Time Unit Lecture Split practical Recording Link Comments
1 1 Jan 11, 13:30 1 (ML Intro) YN - YouTube
2 1 Jan 12, 15:00 1 (ML Intro) YN - YouTube
3 1 Jan 12, 16:00 1 (ML Intro) - BL/SM/YN BL-YouTube/
SM-YouTube/
YN-YouTube
R source/
R solution/
Python Sol
Julia Sol in GitHub
4 1 Jan 13, 17:00 2 (Log Reg NN) BL - YouTube Slides / R example
YN consult 1 Jan 13, 18:00
SM consult 1 Jan 14, 11:00
5 1 Jan 14, 15:00 2 (Log Reg NN) BL - YouTube Slides
6 1 Jan 14, 16:00 2 (Log Reg NN) BL - YouTube Slides
BL consult 1 Jan 15, 16:00
7 1 Jan 15, 17:00 3 (Opt Alg) SM - YouTube
8 2 Jan 18, 17:00 3 (Opt Alg) SM - YouTube
9 2 Jan 18, 18:00 3 (Opt Alg) SM - YouTube
10 2 Jan 19, 12:30 3 (Opt Alg) SM - YouTube
HW 1 2 Due midnight Jan 19 Covers units 1 and 2
11 2 Jan 20, 15:00 4 (General NN) BL - YouTube Slides
12 2 Jan 20, 16:00 4 (General NN) BL - YouTube Approx / Overfit / Nonlinear
YN consult 2 Jan 20, 18:00
SM consult 2 Jan 21, 11:00
13 2 Jan 21, 17:00 4 (General NN) BL - YouTube
14 2 Jan 22, 12:30 4 (General NN) - BL/SM/YN BL-YouTube/
SM-YouTube/
YN-YouTube
R source
BL consult 2 Jan 22, 16:00
Quiz 1 2 Flexible time during Saturday Jan 23 Covering units 2 and 3
15 3 Jan 25, 16:00 5 (Conv NN) YN - YouTube
16 3 Jan 25, 17:00 5 (Conv NN) YN - YouTube
SM consult 3 Jan 27, 11:00
BL consult 3 Jan 27, 16:00
17 3 Jan 27, 17:00 5 (Conv NN) YN - YouTube
18 3 Jan 28, 16:00 5 (Conv NN) - BL/SM/YN BL-YouTube / SM-YouTube / YN-YouTube Keras-R-cheat-sheet / R source
19 3 Jan 28, 17:00 6 (NN Tricks) BL - YouTube Slides
20 3 Jan 29, 17:00 6 (NN Tricks) BL - YouTube
YN consult 3 Jan 29, 19:00
Quiz 2 3 Flexible time during Saturday Jan 30 Covering units 4, 5
21 4 Feb 1, 16:00 7 (GAN) YN - YouTube
22 4 Feb 1, 17:00 7 (GAN) YN - YouTube
HW 2 4 Due midnight Feb 2 Covers units 2, 3, 4, 5
23 4 Feb 2, 17:00 8 (Seq NN) BL - YouTube Slides
SM consult 4 Feb 3, 11:00
24 4 Feb 3, 16:00 8 (Seq NN) BL - YouTube
YN consult 4 Feb 3, 19:00
25 4 Feb 3, 17:00 9 (Deep RL) SM - YouTube
BL consult 4 Feb 4, 16:00
26 4 Feb 4, 17:00 9 (Deep RL) SM - YouTube
27 4 Feb 5, 16:00 9 (Deep RL) SM - YouTube
28 4 Feb 5, 17:00 10 (Apps) YN+SM+BL - YouTube
Quiz 3 4 Flexible time during Saturday Feb 6 Covering units 7,8,9
Project Checkpoint 4 Due midnight Monday Feb 8
SM consult 5 Feb 10, 11:00
YN consult 5 Feb 10, 18:00
BL consult 5 Feb 12, 16:00
Final project Due Feb 15, midnight
  • There is a single Zoom link for the main lectures. There are also three additional zoom links, one per lecturer (BL, SM, and YN). The links are provided via Canvas, accessible to those registered in the course. Use the main link for the main lectures and use the additional links for the split practicals and consultation meetings.
  • BL uses the R software, SM uses Python, and YN uses Julia. In the split practical, join the Zoom link of the lecturer that uses the environment of your choice (choose one out of the three).
  • All lectures and practicals will be recorded and posted online (barring any technical errors).
  • There are three (optional) weekly consultation meetings, one with each lecturer. These will take place in each of the four teaching weeks as well as in the week following Week 4 (allowing help with the project). Consultation meetings are not recorded. A consultation meeting is up to one hour, however join at the start of the meeting if you wish to queue for help/advise, as the meeting will be over once there are no further questions.
  • A lecture hour is always 50 minutes.

Assessment

Here are the assesment items with weight as part of total course grade. Due dates are stated in the schedule above. Late submission is only accepted due to medical reasons or similar.

Assessment is marked by Georgianna Kate Berthaly-Martyn from The University of Queensland.

Handin instructions

All assessment items are to be submitted via the course’s Canvas site. Quiz and assignment submissions are to include a single PDF file. The project is to include more files as specified in the project description. In addition, every submission of an assessment item is to include an accompanying audio clip as specified below.

  • Submit a single PDF file (except for the project), with pages of uniform size, and a file size that does not exceed 8MB (you can use a pdf compression utility if needed). The name of the PDF file should be FFFF_LLLL_HI-IIII.pdf where FFFF is your first name(s), LLLL, is your last name(s), HI abbreviates your home institution(s), and IIII is “Quiz1”, “Assignment1”, etc.

  • Do not submit code or data in a stand alone file (except for the project) - instead format your code into the single PDF file mentioned above. Pixelated (low resolution) screen shots of code and output are not acceptable. Make sure to include your code and output with your submission in a neat manner.

  • Both handwritten notes, typed notes, or a hybird of the two are acceptable. However typed (LaTeX/Jupyter) mathematics is preferable (yet this will not affect your grade). For the project, handwriting is not acceptable. For the quizzes, a scanned/photo image of handwriting is recommended.

  • All graphs, plots, source code, and other figures must be clearly labeled.

  • All questions/items must appear in order as this allows for more effective marking.

  • Every submission includes a recorded audio clip. The audio clip should be in any standard format with a minimum duration of 30 seconds and a maximum duration of 90 seconds. The file size may not exceed 4MB. In your recording, state your name(s), and your experience(s) with this assesment item. Mention the resources that you used to carry out the assignment, and if valid, indicate that you did not plagiarize, stating that the work is only yours. Name the audio clip in the same way that your PDF file is named, but with the valid audio format extension. The same audio clip may be used for all members of a group (in case of group submissions), and may include the voice of a single member speaking on behalf of the group. However, all members must upload the audio clip into Canvas individually (and one member also uploads the PDF file for the whole group).

  • Deviation from these instructions will result in a reduction of your grade.

Student project videos

These are posted with the students’ permission. Here are some videos created by course students.

Page built: 2021-03-04 using R version 4.0.3 (2020-10-10)