Parallelization in your R code, Kaggle and Google team up for a free AI-101, and "Scorigami" interactivity
Check out what's happening in the intersection of sports and data analytics in this week's Data Points of the Week
Mastering Parallelization in R
At Athlyticz, we’ve released a snippet of Dr. Paul Sabin’s course, BreeZing Through the Tidyverse. In this module, Paul discusses parallelization in R, explaining how to distribute tasks across multiple cores of a computer for faster computation. Paul emphasizes the use of packages like doParallel and forEach for parallel execution, particularly useful for computationally intensive tasks like simulations. The lesson demonstrates parallel simulations and showcases code examples for efficient implementation. See the module contents below available to download for free:
Supplemental Materials: Slides, code, and data
Interested in more tutorial content for large-scale sports analytics projects? Could you benefit from a more detailed explanation on a particular topic? Let us know what content you want to see! We’ll be releasing more tutorials in the coming weeks.
Kaggle x Google in Gen AI
Kaggle and Google are teaming up to offer a 5-day Generative AI course from November 11th to 15th. This free course, created by a team of Google’s ML researchers and engineers, is specifically designed to build fundamental knowledge of technologies and techniques behind Gen AI. Although this program isn’t exactly sports related, it offers a unique opportunity for a conceptual deep dive and hands-on coding from top researchers. The course is expected to offer daily assignments, Discord discussion threads, and daily livestream seminars. Follow the curriculum and course topics below:
Day 1: Foundational Models & Prompt Engineering
Day 2: Embeddings and Vector Stores/Databases
Day 3: Generative AI Agents
Day 4: Domain-Specific LLMs
Day 5: MLOps for Generative AI
Our team at Athlyticz will be participating in this course! Follow along with us and learn to apply our newly learned Gen AI knowledge in sports analytics examples!
Slick “Scorigami” Visualizations
Shri Khalpada and Rob Moore, co-founders of PerThirtySix, released an excellent interactive visualization analyzing the NFL phenomenon known as a “scorigami.” A game is said to end in a scorigami if the final score is unique, that is, it has never happened before in the history of the NFL. In the interactive plot, you can see every final NFL score in history with the color field indicating how many times each has happened. The user is also able to slice the data in various ways to adjust the plotting landscape.
Will any of this week’s NFL matchups result in a scorigami? The data (provided by Pro Football Reference) is updated on a weekly basis, so follow along for updates to this fun and creative visualization!