Pinned Tweet
I am very excited that bigergm is available on CRAN.
cran.r-project.org/web/packa…
It is an R package born from a great collaboration with Juan Nelson Martinez Dahbura, Shota Komatsu, Takanori Nishida at Sansan, Inc, Cornelius Fritz and Michael Schweinberger at Penn State University.
Angelo Mele retweeted
1/ Economists: I’ve launched a small project to make AI coding tools actually useful for your day‑to‑day work.
Awesome Econ AI Stuff – a growing collection of reusable AI “skills” for Claude Code, Cursor and Gemini CLI, tailored to empirical + theoretical economics. meleantonio.github.io/awesom…
Angelo Mele retweeted
Modern supply chains don't look like trade theory 101!
They involve constant border crossings, each now hit by tariffs.
Tariffs raise prices, but the more important thing they do is disrupt supply relationships.
1/
graphic h/t @scottlincicome .
This summer Co-Pierre Georg presented our joint work with @cornelius_fritz and Michael Schweinberger on software networks at the 25th ACM Conference on Economics and Computation (EC'24). This is the video of his presentation. youtu.be/-7Xu487kHU0?si=T5Pz…
This version of the paper analyzes 35, 473 repositories with 696, 790 software dependencies of the Rust's ecosystem of libraries. We estimate a structural model of network formation and use it to guide an analysis of the systemic risk embedded in the dependencies.
1
1
59
This risk is highly skewed with the most systemic repository affecting almost 90% of all repositories only two steps away. Protecting only the ten most important repositories reduces vulnerability contagion by nearly 40%. dl.acm.org/doi/10.1145/36708…
1
44
Angelo Mele retweeted
The 2024 Summer Conferences welcomed over 350 economists to Yale’s campus to take part in lively discussion of over 60 new and ongoing research papers.
Read the recap, and check out the papers here: cowles.yale.edu/news/240723/…
My first time at the @TheOfficialACM Conference on Economics and Computation (EC24). I learned a lot on LLMs and AI. Today my coauthor Co-Pierre Georg will present our paper on software dependency networks. All in all a good experience, and hope to come back next year.
We are thrilled to host the 2024 @TheOfficialACM Conference on Economics + Computation (EC'24), featuring 200+ papers that push the frontiers of research in areas like market design, machine learning, LLMs, & information design.
Learn more + full program: economics.yale.edu/news/2407…
13
978
Very excited to present at the @YaleCowles Conference in Econometrics tomorrow. This year's theme is Networks. Great line up and lots of interesting papers!
cowles.yale.edu/events/summe…
Angelo Mele retweeted
I did the same on my very first day with my brother @angelo_mele. We actually said, clearly nobody is using this door while the lateral ones are overcrowded, let's be smart! We did both OK for those worried
Angelo Mele retweeted
I hope all the social network folks are excited about the International Sunbelt Social Networks conference in Edinburgh, Scotland in less than a month! Have you ever tried to analyze large networks with thousands, tens of thousands, or hundreds of thousands of actors?
We have already succesfully used the package in our research and have been able to estimate ERGMs models with community structure with more than 250,000 nodes. You can find some of this work in my webpage. meleangelo.github.io/researc…
1
128
Angelo Mele retweeted
1/4: How often employers include wages in job postings and for which jobs is an important consideration when thinking about how labor markets function. This paper is a simple one: document how often wages are available. Answer: Very little.
Less than 11 percent of online job posts contain wage information, and those which do present wide ranges. Presence and precision of wages vary systematically, from Honey Batra, Amanda Michaud, and @Simon_Mongey nber.org/papers/w31984
Angelo Mele retweeted
If you ran PCA/svd and it localized, it would have this vibe… picking one data point and returning that one. No summarization, no averaging. Just copying a data point in your training data. Gpt does this and so openAI is getting in trouble for copyright infringement.
Very useful method to recover social networks using panel data.
📢 Check out the latest version of “Recovering Social Networks from Panel Data” with the brilliant @ImranRasul3 (@EconUCL) and @pedroclsouza (@QMUL), just accepted at the @RevEconStudies! Here comes a 🧵on what we offer in this paper... 1/6
arxiv.org/abs/1910.07452
