lsstc-dsfp / session-19

This is a repository for Session-19 (Machine Learning) of the LSSTC Data Science Fellowship Program.

Session 19

The nineteenth session of the LSSTC DSFP was hosted by Drexel University in September 2023 and the curriculum covered Machine Learning.

The guest instructors for the S19 were:
Viviana Acquaviva
John Wu
Niharika Sravan
Vicki Toy-Edens

Additional lectures were given by the DSFP leadership team:
Bryan Scott
Adam Miller
Lucianne Walkowicz

Schedule

Day 0 – The Beginning | Introduction for the New Cohort

"The future ain't what it used to be."

~ Yogi Berra

Two orientation lectures are provided asynchronously, these are:

• A Brief Introduction to git/GitHub; B Scott
• Building Visualizations Via Principles of Design ; A Miller

Saturday, Sep 9, 2023

• 10:30 AM - 11:00 AM Registration & Introductions,
• 11:00 AM - 11:30 AM Incoming Student Survey
• 11:30 AM - 12:15 PM Introduction to the Vera C Rubin Observatory and Legacy Survey of Space & Time; L. Walkowicz
• 12:15 PM - 12:30 PM Goals of the DSFP; B. Scott
• 12:30 PM - 01:30 PM LUNCH (provided) & Discussion of the Code of Conduct; B. Scott
• 01:30 PM - 02:45 PM Probability and Data Solutions; A. Miller
• 02:45 PM - 04:00 PM Introduction to Bayesian Statistics Solutions; B. Scott
• 04:00 PM - ??? Break

Day 1 – An Introduction to Machine Learning & Unsupervised Learning

"42."

~ Deep Thought on the answer to life, the universe, and everything (The Hitchhiker's Guide to the Galaxy).

Sunday, Sep 10, 2023

• 09:00 AM – 09:30 AM o Introduction of Cohort 7 and the new instructors
• 09:30 AM – 09:45 AM o Introduction to Hack Sessions
• 09:45 AM – 10:30 AM o Lecture I – Introduction to Machine Learning; B. Scott
• 10:30 AM – 11:00 AM o BREAK
• 11:00 AM – 12:15 PM o Problem | Solutions I – Introduction to ML; B. Scott
• 12:15 PM – 01:45 PM o LUNCH
• 01:45 PM – 02:30 PM o Lecture II – Introduction to Unsupervised Learning; A. Miller
• 02:30 PM – 03:30 PM o Problem II – Introduction to Unsupervised Learning; A. Miller
• 03:30 PM – 03:40 PM o BREAK
• 04:00 PM – 05:00 PM o Lecture III – Introduction to Dimensionality Reduction; B. Scott
• 05:00 PM - 06:00 PM o Problem | Solutions III – Introduction to Dimensionality Reduction; B. Scott

Day 2 – Supervised Machine Learning, Tree, & Ensemble Methods

"I have an infinite capacity for knowledge, and even I'm not sure what is going on outside..."

~GladOS (Portal)

Monday, Sep 11, 2023

• 09:00 AM – 10:30 AM o Lecture IV – Introduction to Supervised Machine Learning; V. Acquaviva
• 10:30 AM – 11:00 AM o BREAK
• 11:00 AM – 12:00 PM o Problem – Introduction to Supervised Machine Learning; V. Acquaviva
• 12:00 PM - 01:30 PM o LUNCH
• 01:30 PM – 02:30 PM o Lecture V – Tree & Ensemble Methods; V. Acquaviva
• 02:30 PM – 03:30 PM o Problem: Tree & Ensemble Methods; V. Acquaviva
• 03:30 PM - 04:00 PM o BREAK
• 04:00 PM - 05:30 PM o Lecture VI – Building Perceptrons for Classification; A. Miller
• 06:00 PM - ??:?? PM o Group dinner

Day 3 — Convolutional Neural Networks

"I am capable of distinguishing over one hundred and fifty simultaneous compositions. But in order to analyze the aesthetics, I try to keep it to ten or less."

~ Lt. Cmdr. Data (Star Trek: The Next Generation)

Tuesday, Sep 12, 2023

• 09:00 AM - 10:00 AM o Lecture VII – Convolutional Neural Networks, J. Wu
• 10:00 AM - 10:30 AM o BREAK
• 10:30 AM - 12:00 PM o Problem: Convolutional Neural Networks J. Wu
• 12:00 PM - ??:?? PM o BREAK

Day 4 — Graph Neural Networks and Reinforcement Learning

"It seems you feel our work is not of benefit to the public."

~ Rachael (Blade Runner)

Wednesday, Sep 13, 2023

• 09:00 AM – 10:00 AM o Lecture VIII – Graph Neural Networks; J. Wu
• 10:00 AM – 10:30 AM o BREAK
• 10:30 AM – 12:00 PM o Problem: Graph Neural Networks; J. Wu
• 12:00 PM – 01:30 PM o LUNCH
• 01:30 PM – 02:30 PM o Lecture IX – Introduction to Reinforcement Learning and The Upper Confidence Bound; A. Sravan
• 02:30 PM – 04:00 PM o Problem: The Upper Confidence Bound; A. Sravan
• 04:00 PM – 04:30 PM o BREAK
• 04:30 PM – 05:00 PM o Hack Pitch Session

Day 5 — Reinforcement Learning (cont.) & Hack Session

"The 9000 series is the most reliable computer ever made. No 9000 computer has ever made a mistake or distorted information. We are all, by any practical definition of the words, foolproof and incapable of error."

~ HAL 9000 (2001: A Space Odyssey)

Thursday, Sep 14, 2023

• 9:00 AM - 10:00 AM o Lecture X – Thompson Sampling; A. Sravan
• 10:00 AM – 10:45 AM o Problem: Thompson Sampling; A. Sravan
• 10:45 AM – 11:00 AM o BREAK
• 11:00 AM – 12:00 PM o Lecture XI – Professional Development: CV Workshop; V. Toy-Edens
• 12:00 PM – 01:00 PM o LUNCH
• 01:00 PM – 04:30 PM o Hack Session;
• 04:30 PM – 05:00 PM o Hack tag–up & Meeting wrap up