A nice plenary talk by @vidalthi

8/8.🔍 Learn more by reading the full paper for details on the methodology and results. 👩‍💻📖

7/8.💻 Curious to try it out? We’ve made our source code publicly available! Check out the code and start detecting infeasible solutions in your own MOBO experiments:

6/8.🚀 This work opens up new possibilities for improving MOBO algorithms. By focusing on infeasibility, researchers can now ensure their algorithms are both effective and reliable in complex optimization scenarios.

5/8.📊 To validate our method, we applied it to benchmark and real-world MOBO problems. The results? Our framework improved the accuracy of performance evaluations, allowing for more reliable comparisons of algorithm efficiency.

4/8.💡 How do we do it? We use data-driven methods, using information from multiple evolutionary algorithms. Our approach identifies infeasibility by analyzing similarities in upper-level decisions and corresponding lower-level objective values.

3/8.🚨 Infeasible solutions can seriously distort performance evaluations. Traditional indicators often ignore the presence of infeasibility, leading to biased comparisons between algorithms. Our framework solves this problem by detecting infeasible solutions. 🔍.

2/8.🌍 MOBO problems are complex! They involve two levels of optimization: the leader and the follower. The leader's decisions directly influence the follower’s response, making the search for optimal solutions incredibly challenging. 🤯 #BilevelOptimization.

🧵1/8.🚀 Excited to share our latest work: "An Efficient Data-Driven Framework for Detecting Infeasible Solutions in Multiobjective Evolutionary Bilevel Optimization (MOBO)." A thread on why detecting infeasible solutions matters and how our framework tackles this! 👇 #AI.

Slides here ⏬.

Slides available after presentation, stay tuned 👀.

Here is the schedule 👉

This Friday, I'll be presenting a tutorial on Metaheuristics.jl, a #julialang package for optimization

Our paper combining bilevel optimization and mechanical design is now published and available in PDF 📄💾. 👉

RT @dotnetschizo: me showing my gf my neovim config

RT @dev_avocado: git commit -m "multiple minor fixes"

RT @ProfFeynman: I was an ordinary person who studied hard. There's no miracle people.

Same in Mexico.

RT @ThePhDPlace: Academia be like.

RT @hackinarticles: Pic of the Day. #infosec #cybersecurity #cybersecuritytips #pentesting #cybersecurityawareness #informationsecurity htt….