#mltheory search results
Mehryar Mohri presents generalization bounds for data dependent hypothesis classes bases on hypothesis set stability, generalizes known Rademacher complexity bounds. #MLTheory
"On the eve of the first Russian revolution, when the Russo-Japanese war had already begun, the #Bolsheviks and the #Mensheviks acted as two separate political groups.” fightbacknews.org/articles/red-r… #MLTheory #Marxism #MarxismLeninism
📢 We are excited to announce our 7th *virtual* mentoring workshop which will be held on November 13 - 14, 2024. Registration (docs.google.com/forms/d/e/1FAI…) is open and free for all. See the full schedule at let-all.com/fall24.html. #MLTheory [1/4]
Which norm should you trust to evaluate generalization🤔? Check out our new paper [arxiv.org/abs/2509.21181] which shows--in closed form--the fascinating scaling law with data for ALL norms between ℓ₁ and ℓ₂. Norm choice ≠ trivial. #MLtheory
![sofuncheung0102's tweet image. Which norm should you trust to evaluate generalization🤔? Check out our new paper [arxiv.org/abs/2509.21181] which shows--in closed form--the fascinating scaling law with data for ALL norms between ℓ₁ and ℓ₂. Norm choice ≠ trivial. #MLtheory](https://pbs.twimg.com/media/G1x5SE1WMAARGeY.jpg)
Why is the Bias-Variance Tradeoff so critical in ML? Master this fundamental concept in just minutes: buff.ly/42vvNQs #MachineLearning #AI #MLTheory
📢For the first time ever Conference on Learning Theory (COLT) 2025 will feature 1 day dedicated to contributed workshops, tutorials, and community-building events. Conference is on June 30-July 4 in Lyon, France. Submit a proposal by April 18 (AOE). #MLTheory #COLT2025
📢 Our third mentoring workshop will be held on Mar 14-15, 2022, in collaboration with the Conference on Algorithmic Learning Theory (ALT 2022). Registration (forms.gle/QBH6FanKjGBTRE…) is open and free for all. See the schedule at let-all.com/alt22.html. #MLTheory
Your model’s stuck at a saddle point, and you don’t know it. Learn the Hessian.📘 Pre-order Mathematics of Machine Learning and follow Packt DataPro on LinkedIn. 👉packt.link/OYc5b #MathForML #AI #MLTheory #100DaysOfMathematicsOfML
📢 Our fourth mentoring workshop will be held on Oct 26-27, 2022, in collaboration with FODSI. Registration (forms.gle/pyF1xKNjXL2UjT…) is open and free for all. See the schedule at let-all.com/fall22.html. #MLTheory [1/4]
Before ChatGPT, there was calculus. The Jacobian powers backprop - math is the original pre-training. 📘 Pre-order Mathematics of Machine Learning + follow Packt DataPro on LinkedIn for more.👉packt.link/OYc5b #MathForML #AI #MLTheory
📢 Our next mentoring workshop will be held on Aug 4-5, 2021, in collaboration with the Conference on Learning Theory (COLT 2021). Registration (forms.gle/i7pTLsWW45bG76…) is open and free for undergrads, grads, and postdocs. See the schedule at let-all.com/colt.html. #MLTheory
Key finding 3: RL’s Razor principle. On-policy RL picks high-reward solutions that stay close to the original model in KL terms. By sampling its own distribution, RL avoids big shifts, reducing forgetting. 🛠️ #ReinforcementLearning #MLTheory 5/8
Design tips from theory Want symbolic behavior? Enforce symmetry, encourage low-entropy weights, and get generalization “for free.” Connects to trends in large-scale reasoning. #MLTheory
📢 We are excited to announce our sixth *virtual* mentoring workshop which will be held on Jun 4-5 2024. Registration (forms.gle/ACZBLto6MweLd9…) is open and free for all. See the full schedule at let-all.com/spring24.html. #MLTheory [1/4]
In Regards the remaining Zodiac Miraculous, if anyone grew up watching the Jackie Chan Adventures cartoon on the weekends as I did, What if the powers were just like the talismans from that show? #MLTheory #MiraculousLadybug
This is awesome! More ML theory conferences should have tutorials and workshops! One of the best ways to learn about new and exciting topics and how they relate to classical questions in the area. COLT and Assoc. for Computational Learning, please? #MLTheory #MachineLearning
Finally, Shay Moran's tutorial on "What Is The Sample Complexity of Differentially Private Learning?" (3/3) youtube.com/watch?v=wk910A…
youtube.com
YouTube
What Is The Sample Complexity of Differentially Private Learning?
👩🏫 Teaching RL in my Grad #MLTheory course and want to shout 📣: RL isn't just another topic, it's the edge of distribution shift! 📈 🤔 Spotted so many #RL misconceptions out there. My own RL journey began at my postdoc time at #MSRNYC 🏙️with @JohnCLangford and Robert…
Kreinovich present mathematical insights showing neural networks can learn from small datasets — and quickly. This hints at overlooked potential in current models. #FewShotLearning #SampleEfficiency #MLTheory
Many architectural choices in #DeepLearning — like activation and pooling functions — have been heuristic or empirical. But what if these choices can be explained through natural mathematical principles? Find out in the “why” part of the talk. #MLTheory #NeuralNetworks
Which norm should you trust to evaluate generalization🤔? Check out our new paper [arxiv.org/abs/2509.21181] which shows--in closed form--the fascinating scaling law with data for ALL norms between ℓ₁ and ℓ₂. Norm choice ≠ trivial. #MLtheory
Key finding 3: RL’s Razor principle. On-policy RL picks high-reward solutions that stay close to the original model in KL terms. By sampling its own distribution, RL avoids big shifts, reducing forgetting. 🛠️ #ReinforcementLearning #MLTheory 5/8
"On the eve of the first Russian revolution, when the Russo-Japanese war had already begun, the #Bolsheviks and the #Mensheviks acted as two separate political groups.” fightbacknews.org/articles/red-r… #MLTheory #Marxism #MarxismLeninism
Today, we break down Reasoning by Superposition, a paper that mathematically models how continuous reasoning unfolds inside AI via superposed embeddings. A leap in understanding AI’s internal logic! 🎧 Watch here: youtu.be/keGOAgVVWaA #AI #MLTheory #LLMs
youtube.com
YouTube
Reasoning by Superposition: Theory Behind AI’s Inner Loop
Design tips from theory Want symbolic behavior? Enforce symmetry, encourage low-entropy weights, and get generalization “for free.” Connects to trends in large-scale reasoning. #MLTheory
Kreinovich present mathematical insights showing neural networks can learn from small datasets — and quickly. This hints at overlooked potential in current models. #FewShotLearning #SampleEfficiency #MLTheory
Many architectural choices in #DeepLearning — like activation and pooling functions — have been heuristic or empirical. But what if these choices can be explained through natural mathematical principles? Find out in the “why” part of the talk. #MLTheory #NeuralNetworks
Your model’s stuck at a saddle point, and you don’t know it. Learn the Hessian.📘 Pre-order Mathematics of Machine Learning and follow Packt DataPro on LinkedIn. 👉packt.link/OYc5b #MathForML #AI #MLTheory #100DaysOfMathematicsOfML
Before ChatGPT, there was calculus. The Jacobian powers backprop - math is the original pre-training. 📘 Pre-order Mathematics of Machine Learning + follow Packt DataPro on LinkedIn for more.👉packt.link/OYc5b #MathForML #AI #MLTheory
📢For the first time ever Conference on Learning Theory (COLT) 2025 will feature 1 day dedicated to contributed workshops, tutorials, and community-building events. Conference is on June 30-July 4 in Lyon, France. Submit a proposal by April 18 (AOE). #MLTheory #COLT2025
Why is the Bias-Variance Tradeoff so critical in ML? Master this fundamental concept in just minutes: buff.ly/42vvNQs #MachineLearning #AI #MLTheory
📢 We are excited to announce our 7th *virtual* mentoring workshop which will be held on November 13 - 14, 2024. Registration (docs.google.com/forms/d/e/1FAI…) is open and free for all. See the full schedule at let-all.com/fall24.html. #MLTheory [1/4]
🧠 Teori Dasar ML: 1. Model: Representasi matematis dari proses yang dipelajari oleh algoritma. 2. Feature: Input variabel yang digunakan model untuk membuat prediksi. 3. Training: Proses belajar untuk menyesuaikan parameter model. 4. Validation: Mengukur kinerja model #MLTheory
📢 We are excited to announce our sixth *virtual* mentoring workshop which will be held on Jun 4-5 2024. Registration (forms.gle/ACZBLto6MweLd9…) is open and free for all. See the full schedule at let-all.com/spring24.html. #MLTheory [1/4]
New paper develops simplified theory for regression algorithms, explaining their success in finding patterns amid noise. Results help demystify 'black box' methods. #MLtheory #regression #analytics Read more: bulletpape.rs/YXNip7r
New theory shows kernel ridge regression optimal for low-complexity problems, but hits limits in high dimensions. Convergence rates evolve in unexpected waves as data size varies. #MLTheory #KRR #Generalization Read more: bulletpape.rs/fUGrfaD
When you have been jumping around thinking about new settings for so long and you realize that you finally have something to prove again. #MLTheory #MachineLearning
👩🏫 Teaching RL in my Grad #MLTheory course and want to shout 📣: RL isn't just another topic, it's the edge of distribution shift! 📈 🤔 Spotted so many #RL misconceptions out there. My own RL journey began at my postdoc time at #MSRNYC 🏙️with @JohnCLangford and Robert…
Mehryar Mohri presents generalization bounds for data dependent hypothesis classes bases on hypothesis set stability, generalizes known Rademacher complexity bounds. #MLTheory
Why is the Bias-Variance Tradeoff so critical in ML? Master this fundamental concept in just minutes: buff.ly/42vvNQs #MachineLearning #AI #MLTheory
Which norm should you trust to evaluate generalization🤔? Check out our new paper [arxiv.org/abs/2509.21181] which shows--in closed form--the fascinating scaling law with data for ALL norms between ℓ₁ and ℓ₂. Norm choice ≠ trivial. #MLtheory
New article in our series on #MLTheory: It is historical materialism that demonstrates the link between dialectical materialism and political economy. Here we have dialectical materialism applied to history. #marxism #marxismleninism fightbacknews.org/2022/5/9/what-…
Your model’s stuck at a saddle point, and you don’t know it. Learn the Hessian.📘 Pre-order Mathematics of Machine Learning and follow Packt DataPro on LinkedIn. 👉packt.link/OYc5b #MathForML #AI #MLTheory #100DaysOfMathematicsOfML
Before ChatGPT, there was calculus. The Jacobian powers backprop - math is the original pre-training. 📘 Pre-order Mathematics of Machine Learning + follow Packt DataPro on LinkedIn for more.👉packt.link/OYc5b #MathForML #AI #MLTheory
"On the eve of the first Russian revolution, when the Russo-Japanese war had already begun, the #Bolsheviks and the #Mensheviks acted as two separate political groups.” fightbacknews.org/articles/red-r… #MLTheory #Marxism #MarxismLeninism
Something went wrong.
Something went wrong.
United States Trends
- 1. Rams 22.5K posts
- 2. Seahawks 28.1K posts
- 3. 49ers 19.7K posts
- 4. Lions 74.4K posts
- 5. Commanders 81.9K posts
- 6. Niners 4,746 posts
- 7. Stafford 7,657 posts
- 8. Canada Dry 1,116 posts
- 9. Bills 143K posts
- 10. Cardinals 10.4K posts
- 11. #OnePride 4,278 posts
- 12. Dan Campbell 2,726 posts
- 13. Giants 71.1K posts
- 14. #RaiseHail 3,285 posts
- 15. DO NOT CAVE 9,954 posts
- 16. Daboll 14.5K posts
- 17. Gibbs 7,307 posts
- 18. Jags 7,561 posts
- 19. Dart 28.4K posts
- 20. Bears 63.8K posts