A Hands-On Introduction To Multi-Objective Optimization
summary
Hands-on material uses Git-hosted Jupyter notebooks and Google Colab to solve a knapsack: minimize bag weight while maximizing value, with code that finds Pareto Fronts and traces generations. An evolutionary genetic algorithm (package named deep in the materials) explores a huge decision space (≈10^26 in the protein case) and yields a frontier of trade-off solutions; applied to antibody design it outputs ~96 candidate sequences and in a case study reduces false positives by ~8% and misclassification by ~5%.
AI-written summary · about
chapters (29)
- 0:00Introduction
- 0:35MultiObjective Optimization
- 2:32Objectives
- 4:41About Me
- 6:17Agenda Parts
- 7:17Welcome Motivation
- 7:28Play Button
- 8:07Google Collab
- 9:02Summary
- 9:16Part 1 Challenge
- 10:23Part 2 Challenge
- 11:22Quick Summary
- 12:41Application
- 15:34Deep
- 17:16Notebook
- 18:07Notebook Setup
- 18:49Deep Setup
- 20:47Custom Functions
- 21:28Knapsack Problem
- 21:57Visualization
- 22:07Final Result
- 22:17Hack
- 22:28Motivation
- 23:20Case Studies
- 24:22Case Study 1
- 26:41Decision Space Animation
- 27:36Data Science Example
- 29:19Machine Learning Example
- 31:50Summarize
related talks
34mImproved Decision Making with Pareto Fronts
01h 28mFollow the gradient: an introduction to mathematical optimisation
30mEvolutionary Algorithms Perfecting the Art of "Good Enough"
46mTight Genes: Intro to Genetic Algorithms
10mMulti-objective Optimization for COVID Vaccine Management
48mTight Genes: Intro to Genetic Algorithms
50mOptimal control, no! Than what? (...)
01h 15mEducational framework for Black Box optimization methods design