Prof. dr. Danny Dig, University of Colorado: Learning and Applying Best Practices for Code Evolution

Vă invităm sa participați la workshopul organizat în cadrul grantului ORCHESTRA.

Assoc. Prof. Dr. Danny Dig va susține prezentarea:

Learning and Applying Best Practices for Code Evolution

Invitat: Prof. dr. Danny Dig, University of Colorado, Computer Science Department, http://dig.cs.illinois.edu/

Cand?
Joi, 10 November 2022
17:30 – 18:30 (EET)

Unde?
Online: Teams meeting

Vă asteptam!
Echipa ORCHESTRA

https://www.cs.ubbcluj.ro/~avescan/orchestra

Title: Learning and Applying Best Practices for Code Evolution

Abstract: Our community is at a crossroads now. Despite the extraordinary rise in popularity of Python-based ML systems, they do not benefit from the advances in the software evolution field. The science and tooling for developing and evolving ML code-bases is significantly behind. To reduce this gap, we conducted the first and most fine-grained study on code change patterns in a diverse corpus of 1000 top-rated ML systems.  We triangulated our findings with surveys of hundreds of ML engineers and reveal four major trends of how ML engineers evolve their code. Then we designed, implemented and evaluated a new generation of Data-driven code transformation tools that transplant the best practices for code evolution into new projects. More than 90% of the patches that we submitted to famous open-source projects like TensorFlow and PyTorch were accepted, showing the usefulness and applicability of our tools.

We launch a call to action for researchers, tool builders, ML library vendors, developers, and educators. This will inspire and equip you so that you can make a difference, with people who make a difference, at a time when it makes a difference. I will conclude with the lessons we are learning from working with industry, and invite colleagues  in academia and students to join forces to make a lasting impact through   our recently-launched Center on Intelligence for IoT systems: http://ppicenter.org

Biography: Danny Dig is professor in CS at University of Colorado at Boulder, and an adjunct professor at University of Illinois at Urbana-Champaign.

He enjoys research in Software Engineering, with a focus on interactive program analysis and transformations that improve programmer productivity and software quality. He successfully pioneered the field of refactoring in cutting-edge domains including ML, mobile, concurrency and parallelism, component-based, testing, and end-user programming. He earned his Ph.D. from the University of Illinois at Urbana-Champaign where his research won the best Ph.D. dissertation award, and the First Prize at the ACM Student Research Competition Grand Finals. He did a postdoc at MIT.

He (co-)authored 60+ journal and conference papers that appeared in top places in SE/PL. His group’s research was recognized with 9 paper awards at the flagship conferences in SE, 4 award runner-ups, and 1 most influential paper award. He received the NSF CAREER award, the Google Faculty Research Award (twice), and the Microsoft SE Innovation Award (twice). With his students, they released dozens of software systems. Some of these are shipping with the official release of the popular Eclipse, NetBeans, Android Studio, and Visual Studio development environments which are used daily by millions of developers. His group’s research is funded by NSF, Boeing, IBM, Intel, Google, Microsoft, NEC, and Trimble. He travels all over the world to inspire and equip the next generation of technical leaders.