Declarative Programming for Statistical ML
summary
RELOOP is a declarative mathematical programming language embedded in Python that lets users write convex optimization problems before knowing domain individuals and therefore before variables and constraints exist. It compiles relational SVM and other quadratic programs to exploit symmetry detection and algebraic decision diagrams, uses matrix free optimization to avoid dense factorization, and can make a linear SVM network‑aware with three lines of code. The system scales to tens of millions of edges and runs large iterations on a desktop.
AI-written summary · about
chapters (7)
- 0:00Start
- 4:23Take your spreadsheet ...
- 7:24What are we missing ?
- 10:15Standard ML approach: Support Vector Machines
- 11:00Statistical Machine Learning via Declarative Programming Write down the problem in paper form The machine then compiles automatically into algebraic solver form.
- 12:05Captures the essence of a problem and can be reused for several problems
- 17:36DEMOCRATIZATION OF MACHINE LEARNING
related talks
39mPractical Optimization for Stats Nerds
39mLinear Logic Programming
15mScaling Optimization Over Uncertainty via Compilation
41mScalable Collective Inference from Richly Structured Data using Probabilistic Soft Logic (PSL)
38mProbabilistic programming and meta-programming in Clojure
25mVerl: A Flexible and Efficient RL Framework for LLMs
19mSTRADS-AP: Simplifying Distributed Machine Learning Programming without Introducing a New Programming Model
01h 00mBetter Learning through Programming Languages