_jmejia Profile Banner
Jesús-Adolfo Mejía Profile
Jesús-Adolfo Mejía

@_jmejia

Followers
153
Following
2K
Media
36
Statuses
392

Bilevel Optimization, IA, #JuliaLang, and Cats

México
Joined August 2019
Don't wanna be here? Send us removal request.
@_jmejia
Jesús-Adolfo Mejía
9 months
A nice plenary talk by @vidalthi
Tweet media one
0
0
3
@_jmejia
Jesús-Adolfo Mejía
10 months
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:
Tweet card summary image
github.com
An Efficient Data-Driven Framework for Detecting Infeasible Solutions in Multiobjective Evolutionary Bilevel Optimization - jmejia8/mobo-infeasibility-detector
1
0
2
@_jmejia
Jesús-Adolfo Mejía
10 months
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.
1
0
2
@_jmejia
Jesús-Adolfo Mejía
10 months
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.
1
0
2
@_jmejia
Jesús-Adolfo Mejía
10 months
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.
1
0
2
@_jmejia
Jesús-Adolfo Mejía
10 months
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. 🔍.
1
0
2
@_jmejia
Jesús-Adolfo Mejía
10 months
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
0
2
@_jmejia
Jesús-Adolfo Mejía
10 months
🧵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.
1
1
4
@_jmejia
Jesús-Adolfo Mejía
1 year
Slides here ⏬.
0
0
1
@_jmejia
Jesús-Adolfo Mejía
1 year
Slides available after presentation, stay tuned 👀.
Tweet media one
1
1
7
@_jmejia
Jesús-Adolfo Mejía
1 year
Here is the schedule 👉
Tweet media one
1
0
3
@_jmejia
Jesús-Adolfo Mejía
1 year
This Friday, I'll be presenting a tutorial on Metaheuristics.jl, a #julialang package for optimization
Tweet media one
3
8
29
@_jmejia
Jesús-Adolfo Mejía
1 year
Our paper combining bilevel optimization and mechanical design is now published and available in PDF 📄💾. 👉
Tweet media one
0
1
4
@_jmejia
Jesús-Adolfo Mejía
1 year
RT @dotnetschizo: me showing my gf my neovim config
Tweet media one
0
266
0
@_jmejia
Jesús-Adolfo Mejía
2 years
RT @dev_avocado: git commit -m "multiple minor fixes"
Tweet media one
0
466
0
@_jmejia
Jesús-Adolfo Mejía
2 years
RT @ProfFeynman: I was an ordinary person who studied hard. There's no miracle people.
0
2K
0
@_jmejia
Jesús-Adolfo Mejía
2 years
Same in Mexico.
@docmilanfar
Peyman Milanfar
2 years
the most common cause of traffic accidents in Boston is two cars trying to hit the same pedestrian.
0
0
0
@_jmejia
Jesús-Adolfo Mejía
2 years
RT @ThePhDPlace: Academia be like.
0
786
0