Warren Powell’s Post

Did you know that the "Linear Programming" was originally referred to as "Programming in a Linear Structure"? Did you also know that the term "primal" was actually coined by George Dantzig's father Tobias Dantzig in 1954? If you are interested in exploring the origins of these and other Optimization terms such as "simplex method" and "mathematical programming", then check out this excerpt written by Dantzig himself for the book "History of Mathematical Programming: A Collection of Personal Reminiscences", published in 1991. The article also appeared in 2002 in Operations Research and it can be accessed here: https://lnkd.in/dPEkMK34. Other fascinating stories related to linear programming and mathematical optimization can be found in several episodes of the "Subject to" (s.t.) podcast, including those with legends from the field such as Robert Bixby (https://lnkd.in/dNGeXgMs), Vasek Chvatal (https://lnkd.in/dHYge6S3), and Dimitris Bertsimas (https://lnkd.in/dUNa-d8n). Be sure to check them out!

There are thick lines between Algorithm, Programming and Machine Learning Model--- in plain English.

'Foundations of Probabilistic Logic Programming, 2nd Edition' is now available to purchase! Since its birth, the field of Probabilistic Logic Programming has seen a steady increase of activity, with many proposals for languages and algorithms for inference and learning. This book aims at providing an overview of the field with a special emphasis on languages under the Distribution Semantics, one of the most influential approaches. The book presents the main ideas for semantics, inference, and learning and highlights connections between the methods. https://lnkd.in/ezMxKA3K #RiverPublishers #logic #programming #ai #machinelearning

Learning to program with large-language models For example, ChatGPT or Copilot. Episode 582 of IEEE's Software Engineering Radio discusses the use of LLMs in programming and some consequences. They mention examples of getting quicker samples of code, or even a wider range of samples. Their focus is on students, so they discuss concerns of evaluating samples for correctness or other attributes, but they acknowledge the exposure to a variety of programming styles. They do mention potential use by software professionals. Approx 1 hr podcast. https://lnkd.in/gcwm52TB

𝐃𝐢𝐬𝐭𝐫𝐢𝐛𝐮𝐭𝐢𝐨𝐧𝐚𝐥 𝐑𝐞𝐢𝐧𝐟𝐨𝐫𝐜𝐞𝐦𝐞𝐧𝐭 𝐋𝐞𝐚𝐫𝐧𝐢𝐧𝐠 - a great book about #reinforcementlearning by Marc G. Bellemare, Will Dabney and Mark Rowland 𝕿𝖆𝖇𝖑𝖊 𝖔𝖋 𝕮𝖔𝖓𝖙𝖊𝖓𝖙𝖘: 🎯 Introduction 🎯 The Distribution of Returns 🎯 Learning the Return Distribution 🎯 Operators and Metrics 🎯 Distributional Dynamic Programming 🎯 Incremental Algorithms 🎯 Control 🎯 Statistical Functionals 🎯 Linear Function Approximation 🎯 Deep Reinforcement Learning 🎯 Two Applications and a Conclusion 𝕱𝖗𝖊𝖊 𝕯𝖔𝖜𝖓𝖑𝖔𝖆𝖉 𝕷𝖎𝖓𝖐: https://lnkd.in/dpS2unas

Dive into the world of mathematical thinking! Explore the Pigeonhole Principle with me and discover how this simple concept revolutionizes problem-solving in everyday life and programming. 🚀 Read the article here: https://lnkd.in/gZZwJfHu #Mathematics #ProblemSolving #PigeonholePrinciple

Practical Julia: A Hands-On Introduction for Scientific Minds The Julia programming language is acclaimed in scientific circles for its unparalleled ease, interactivity, and speed. Practical Julia is a comprehensive introduction to the language, making it accessible even if you’re new to programming. Dive in with a thorough guide to Julia’s syntax, data types, and best practices, then transition to craft solutions for challenges in physics, statistics, biology, mathematics, scientific machine learning, and more. Whether you’re solving computational problems, visualizing data, writing simulations, or developing specialized tools, Practical Julia will show you how. As you work through the book, you’ll: Use comprehensions and generators, higher-level functions, array initialization and manipulation, and perform operations on Unicode text Create new syntax and generate code with metaprogramming and macros, and control the error system to manipulate program execution Visualize everything from mathematical constructs and experimental designs to algorithm flowcharts Elevate performance using Julia’s unique type system with multiple dispatch Delve into scientific packages tailored for diverse fields like fluid dynamics, agent-based modeling, and image processing Whether your interest is in scientific research, statistics, mathematics, or just the fun of programming with Julia, Practical Julia will have you writing high-performance code that can do real work in no time. Online Resources: Ready-to-run code samples, illustrations, and supplemental animations available at https://lnkd.in/dTDeaUym. https://lnkd.in/da2twtAu

Top R Programming Project Ideas for Beginners to Try in 2024 https://lnkd.in/gu_BisDJ #RProgrammingProjectIdeasforBeginners #ProgrammingLanguage #DataAnalysis #AI #AINews #AnalyticsInsight #AnalyticsInsightMagazine

🚨🚨 𝐓𝐡𝐞 𝐄𝐥𝐞𝐦𝐞𝐧𝐭𝐬 𝐨𝐟 𝐃𝐢𝐟𝐟𝐞𝐫𝐞𝐧𝐭𝐢𝐚𝐛𝐥𝐞 𝐏𝐫𝐨𝐠𝐫𝐚𝐦𝐦𝐢𝐧𝐠🚨🚨 - a 383-page book about on a fascinating topic. 🔥 𝕿𝖆𝖇𝖑𝖊 𝖔𝖋 𝕮𝖔𝖓𝖙𝖊𝖓𝖙𝖘: ✅ Differentiation ✅ Probabilistic learning ✅ Parameterized programs ✅ Control flows ✅ Finite differences ✅ Automatic differentiation ✅ Second-order automatic differentiation ✅ Inference in graphical models as differentiation ✅ Differentiating through optimization ✅ Differentiating through integration ✅ Smoothing by optimization ✅ Smoothing by integration ✅ Optimization basics ✅ First-order optimization ✅ Second-order optimization ✅ Duality 𝕱𝖗𝖊𝖊 𝕯𝖔𝖜𝖓𝖑𝖔𝖆𝖉:  https://lnkd.in/eN3WfUGn The Elements of Differentiable Programming, Mathieu Blondel, Vincent Roulet; 383 pages. Foundational contribution in deep analysis of what gradient computation really means algorithmically. This contribution is filling a huge gap in the current ML literature. Computer Science > Machine Learning arXiv:2403.14606 (cs) [Submitted on 21 Mar 2024] The Elements of Differentiable Programming Mathieu Blondel, Vincent Roulet Download PDF Artificial intelligence has recently experienced remarkable advances, fueled by large models, vast datasets, accelerated hardware, and, last but not least, the transformative power of differentiable programming. This new programming paradigm enables end-to-end differentiation of complex computer programs (including those with control flows and data structures), making gradient-based optimization of program parameters possible. As an emerging paradigm, differentiable programming builds upon several areas of computer science and applied mathematics, including automatic differentiation, graphical models, optimization and statistics. This book presents a comprehensive review of the fundamental concepts useful for differentiable programming. We adopt two main perspectives, that of optimization and that of probability, with clear analogies between the two. Differentiable programming is not merely the differentiation of programs, but also the thoughtful design of programs intended for differentiation. By making programs differentiable, we inherently introduce probability distributions over their execution, providing a means to quantify the uncertainty associated with program outputs. Submission history From: Mathieu Blondel [view email] [v1] Thu, 21 Mar 2024 17:55:16 UTC (1,921 KB) 🚨🚨 可微分编程的要素🚨🚨 - 一本383页的书，讲述令人着迷的话题。 🔥 目录： ✅ 微分 ✅ 概率学习 ✅ 参数化程序 ✅ 控制流 ✅ 有限差分 ✅ 自动微分 ✅ 二阶自动微分 ✅ 图形模型中的推理作为微分 ✅ 通过优化进行微分 ✅ 通过积分进行微分 ✅ 通过优化进行平滑 ✅ 通过积分进行平滑 ✅ 优化基础 ✅ 一阶优化 ✅ 二阶优化 ✅ 对偶性 Share & Translate: Chinou Gea (秦陇纪) @2024 DSS-SDC, IFS-AHSC. Data Simplicity Community https://lnkd.in/g3TPCuPn #DataSimp #DataScience #computing #program #AI #ArtificialIntelligence #PatternRecognition #DifferentiableProgramming #MachineLearning #ML #mathematics #differentiation #integration

From my training in Mathematics I have a strong preference for functional styles of programming. However, the functional paradigm struggles to represent persistence, such as in-place memory operations, or asynchronous execution. In these types of use cases I find that object oriented styles offer a cleaner paradigm for encapsulating persistence. My rule of thumb is: I like functional programming for: - immediate execution - producer-subscriber flows - type coercion - abstracting the guts of mathematical algorithms. - anything the looks roughly like a pipeline. - stateless processing. I like object oriented programming for: - holding references to asynchronous execution. - persistent connections to unmanaged resources (IO, etc...). - in-place and self modifying algorithms. - anything that looks roughly like an automaton, where you make requests to change it's ongoing behaviour. - stateful programming.