ML ain’t your only hammer: adding mathematical optimisation to the data scientist’s toolbox
summary
Claims that mathematical optimisation complements ML and then shows it by formulating a model in Python, solving a knapsack example and handing the equations to a solver described as a compiler for math (open‑source and commercial solvers). Demonstrates grid dispatch with thousands of bids and constraints, battery state‑of‑charge rules, scenario runs over a 20‑year horizon, and interpretable dual values that set prices including a price swing -$1,000 to $177,000 and produces optimal, auditable decisions.
AI-written summary · about
chapters (10)
- 0:00Mathematical optimization intro
- 1:52Limitations of traditional machine learning
- 3:39Optimization in constrained systems
- 6:46Long-term planning
- 8:09Knapsack problem example
- 12:23Knapsack demo takeaways
- 14:49Advantages over RL
- 15:51Optimization complements ML
- 16:12Finding optimization opportunities
- 18:58Closing and exploration
related talks
01h 28mFollow the gradient: an introduction to mathematical optimisation
39mPractical Optimization for Stats Nerds
18mProduct Management for Optimization
38mFusing Machine Learning and Mixed Integer Linear Programming
01h 15mEducational framework for Black Box optimization methods design
31mBuilding an electricity market model from scratch
16mUsing AI to solve complex business problems: optimizing...
34mForget about AI and do mathematical modeling instead!