Associate Professor @HECLausanneEcon
- Tweets 701
- Following 827
- Followers 1,940
- Likes 3,418
Pinned Tweet
Thrilled to share a project I've been refining: a complete, open-source repository on "Deep Learning for Solving and Estimating Dynamic Models in Economics and Finance."
I've cleaned up the materials from my PhD classes and summer schools into one coherent resource. 🧵 1/6
2
138
617
51,217
Delighted to see this published with @PhilippJRenner!
Moving beyond 2 or 3 types has been a persistent roadblock for ages. By bypassing APS set-valued iterations entirely, we open up rich type distributions for applied work.
💻 Code: github.com/GaussianProcesses…
"Machine Learning for Dynamic Incentive Problems": @comp_simon & @PhilippJRenner introduce a framework combining penalties, Gaussian processes and Bayesian active learning to bypass set-valued dynamic programming. Handles at least 10 persistent types."
restud.com/machine-learning-…
4
21
1,123
Simon Scheidegger retweeted
#Kernenergie hat im #Hitzesommer absolut konstant und zuverlässig Strom produziert. Zusammen mit der #Wasserkraft ist sie das Rückgrat der kontinuierlichen #Stromversorgung.
#AKW
13
22
84
1,928
Simon Scheidegger retweeted
I remember the first presentation of this. @comp_simon drew Gaussian Processes by hand.
Thrilled to share that "Machine Learning for Dynamic Incentive Problems" (with @PhilippJRenner ) is in Review of Economic Studies (@RestudJournal)! This was truly a career-decisive project for me: my Job Market Paper back when ML and deep learning were not yet a thing 🧵👇
2
24
1,458
A huge thank you to the editor @SorryToBeKurt, referees, and colleagues who supported this journey!
Free-access article link: academic.oup.com/restud/adva…
Python toolbox & replication code: github.com/GaussianProcesses…
2
11
5,222
Free-access article link: academic.oup.com/restud/adva…
Python toolbox & replication code: github.com/GaussianProcesses…
We introduce a penalty relaxation that bypasses computing feasible sets a priori, solving it via value function iteration with Gaussian Process regression and Bayesian Active Learning. The generic framework scales up to at least 10 persistent types.
1
8
2,477
We introduce a penalty relaxation that bypasses computing feasible sets a priori, solving it via value function iteration with Gaussian Process regression and Bayesian Active Learning. The generic framework scales up to at least 10 persistent types.
Dynamic adverse selection with persistent private info typically requires set-valued dynamic programming over high-dimensional, unknown, and irregular state spaces (curse of dimensionality).
1
6
1,454
Dynamic adverse selection with persistent private info typically requires set-valued dynamic programming over high-dimensional, unknown, and irregular state spaces (curse of dimensionality).
Thrilled to share that "Machine Learning for Dynamic Incentive Problems" (with @PhilippJRenner ) is in Review of Economic Studies (@RestudJournal)! This was truly a career-decisive project for me: my Job Market Paper back when ML and deep learning were not yet a thing 🧵👇
11
46
3,994
Thrilled to share that "Machine Learning for Dynamic Incentive Problems" (with @PhilippJRenner ) is in Review of Economic Studies (@RestudJournal)! This was truly a career-decisive project for me: my Job Market Paper back when ML and deep learning were not yet a thing 🧵👇
4
20
105
12,897
Simon Scheidegger retweeted
🎓Reminder: PhD Guest Course at #DSEA
@UniPadova
📅23-24 September 2026
'Deep Learning for Solving Dynamic Models' with @comp_simon from @heclausanne.
Topics include ML, DEQNs & macroeconomic applications.
Register now unipd.link/PhD_Guest_Courses… #EconX #EconTwitter
4
13
2,737
Simon Scheidegger retweeted
Cool event @EconomicsUnito and @CollegioCA next week: deep learning summer school with @comp_simon and friends sites.google.com/carloalbert…, round table discussion on thursday
Can AI reshape the future of finance research?
Join leading international scholars at @CCA_Collegio to discuss AI's impact on economics, finance, and academic research.
📅 Aug 27 | 🕔 5:15 PM
🎟️ Register: l1nq.com/w56204o
#AI #Economics #Finance #LLMs #Research #CCA
2
2
10
914
Thrilled that "Deep Learning for Solving and Estimating Dynamic Models in Economics and Finance," found a home at @QuantEcon. Access the e-book/code here: quantecon.org/books/ . Huge thanks to Matt McKay, @john_stachurski , Tom Sargent, QuantEcon team for making this happen!
2
47
192
17,743
I am glad to announce that our paper “Taming the Curse of Dimensionality:
Quantitative Economics with Deep Learning” with @JesusFerna7026 and @jlperla has been officially accepted for publication at @qe_editors
galonuno.com/uploads/1/3/4/6…
6
27
178
19,040
👇very happy and proud to see this in print. There are plenty of accompanying codes that should lower the bar for interested researchers to enter this exciting field.
In the new @EJ_RES Special Issue on Climate Change and Inequality, ‘Building Interpretable Climate Emulators for Economics’ by @aryanEfy, Doris Folini, @friedlale, Felix Kübler, @comp_simon, Olaf Schenk doi.org/10.1093/ej/ueaf131 @RoyalEconSoc #EconTwitter
1
7
28
2,984
Simon Scheidegger retweeted
Das zweitstärkste AKW der USA steht in der Wüste.
Dort soll es recht trocken sein.
en.wikipedia.org/wiki/Palo_V…
9
30
203
3,328
👇very cool work!
So: roughly 60% use LLMs in research. The amount of published papers is skyrocketing (6% now, many more in the pipeline of journals). We looked at reproducibility properties of LLMs in a new paper out today arxiv.org/abs/2607.24372 . tl;dr: LLMs are non-deterministic. Careful! 1/n
1
1
8
3,969
Simon Scheidegger retweeted
🎓PhD Guest Course at #DSEA @UniPadova
📅23-24 September 2026
'Deep Learning for Solving Dynamic Models' with Simon Scheidegger @comp_simon from @heclausanne.
ML, DEQNs & macroeconomic applications.
Info & registration: unipd.link/PhD_Guest_Courses…
5
34
2,310
Very happy see our paper (@friedlale @aryan_eftekhari ) in this special issue:
academic.oup.com/ej/article-…
Thank you @EJ_RES and @albertobisin to put focus on such an important topic. 👇
This is my first baby as a Special Issue Editor at the Economic Journal @EJ_RES. Honestly, I am proud of this. Please check it out.
My very brief Intro and - most importantly - the leading paper by @brunoconteleite, @klausvanieper , @HansbergRossi are free to see.
1
5
22
4,387
It was an honor to give the opening keynote at CEF in Venice. Based on requests, here are my slides. I tried to provide a broad overview of a rapidly evolving, super exciting field. 👇
drive.google.com/file/d/1z0l…
#CEF2026 #Macroeconomics #DeepLearning
1
11
64
4,722
Simon Scheidegger retweeted
Exiting day, starting with the plenary of @comp_simon @unil
If you are attending the CEF conference tomorrow, and still have time on your schedule, here are two things I look forward:
1) rise and shine! 9:00am, Aula magna, keynote on "(Un)masking Neural Networks. for Dynamic Economic Models" by the one and only @comp_simon @unil
4
22
6,628