From eafbfb10d780336df8aa1fc4580c2c03bb53632f Mon Sep 17 00:00:00 2001 From: ruthl Date: Thu, 30 Jan 2025 13:37:08 +1030 Subject: [PATCH] Markdown timestamps for MathOpt Solver video --- ...Independent_Modeling_in_Googles_ORTools.md | 30 +++++++++++++++++++ 1 file changed, 30 insertions(+) create mode 100644 videos/Mathopt_Solver_Independent_Modeling_in_Googles_ORTools.md diff --git a/videos/Mathopt_Solver_Independent_Modeling_in_Googles_ORTools.md b/videos/Mathopt_Solver_Independent_Modeling_in_Googles_ORTools.md new file mode 100644 index 0000000..a7f38e9 --- /dev/null +++ b/videos/Mathopt_Solver_Independent_Modeling_in_Googles_ORTools.md @@ -0,0 +1,30 @@ +References: +Video title: MathOpt: Solver Independent Modeling in Google's OR-Tools | Ross Anderson | JuliaCon 2023 +Video link: https://www.youtube.com/watch?v=L5b4YQowXBg +Speaker: Ross Anderson + +Timestamps: +00:00 Welcome and speaker introduction +01:03 What is MathOpt +02:30 MathOpt Python MIP Example +03:45 MathOpt contributors +04:42 Presentation Outline +05:00 Applications of Mathematical Optimization at Google +07:00 Why is optimization at Google different? +09:50 MathOpt Design and Implementation +12:37 MathOpt High Level Architecture +14:15 Protocol Buffers for Optimization Models +15:53 Protos: Incremental changes in the MathOpt Protocol +16:57 Protos: Strong, Programmatically Enforced Contracts +18:43 Client Library Responsibilities +19:17 Storing sparse matrices +20:30 Efficiently building model updates (skipped due to time constraints) +20:40 Variables and Expressions +21:41 Expression Gotchas +23:55 Solver implementation (skipped due to time constraints) +24:08 Performance Benchmark: TSP +25:55 Demo: Portable C++ (skipped due to internet connection) +26:20 Future Directions for MathOpt +27:23 Conclusion: MathOpt goals: Reliable, Powerful, Maintainable +28:20 Question: Serializing protocol buffers +29:10 Question: Problem size and problem decomposition