'neural networks' 검색 분석 결과

BREAKING: MIT just mass released their Al library for free. Here's the full list of books: Foundations 1. Foundations of Machine Learning Core algorithms explained. Theory meets practice. 2. Understanding Deep Learning Neural networks demystified. Visual explanations included. 3. Machine Learning Systems Production-ready architecture. System design principles. Advanced Techniques 4. Algorithms for ML Computational thinking simplified. Decision-making frameworks.

⚡Physics-Informed Neural Networks (PINN)⚡ PINNs are neural networks trained to solve supervised learning tasks while respecting any given laws of physics described by general nonlinear partial differential equations. The physics-informed neural network can predict the solution far from the experimental data points and thus performs much better than the naive network. This network indeed has some concepts of our prior physical principles.

Yeah, neural networks look cool on your resume and may make your thesis stand out. But there are plenty of simple models that can solve real business problems much more efficiently. Top 8 machine learning algorithms that actually solve business problems

An LSE paper just showed that five neural networks trained to hedge options beat Black-Scholes on tail risk by 41 basis points. The popular read is that neural nets are better at pricing. They aren't. They trade four times less. The entire edge is laziness. The LSTM independently discovered what Whalley and Wilmott derived analytically in 1997: sometimes the best trade is no trade. And that's the least interesting finding in the paper. Here's what actually matters:

I've been coding neural networks for 2 years. Ask me anything about getting started with AI/ML! 🧠

Mathematics of Neural Networks by Bart M. N. Smets PDF: arxiv.org/pdf/2403.04807 📕 400+ 𝗗𝗮𝘁𝗮 𝗦𝗰𝗶𝗲𝗻𝗰𝗲 𝗥𝗲𝘀𝗼𝘂𝗿𝗰𝗲𝘀: https://topmate.io/arif_alam/787013

Currently diving deep into Physics-Informed Neural Networks (PINNs). Exploring scientific ML at the intersection of PDEs + deep learning. If you're serious about research in this space, I’d love to exchange ideas. #STEM #Research #ResearchPapers #MachineLearning

When people say ‘I’m not anti-AI, I’m anti-LLMs,’ I’d gently ask: do you know what architecture AlphaFold uses? It’s neural networks built on attention mechanisms, which is the same foundational research that powers the chatbot you’re dismissing. You’ve drawn a line between things you respect and things you don’t, and the line isn’t technical, it’s social. We do not have good and bad technologies, only good and bad use cases.

The Mathematical Engine of Artificial Intelligence: Understanding Backpropagation Everyone has used AI by now in some form or another. But how does it work. Here's the fundamental breakdown. At the heart of modern machine learning lies a sophisticated optimisation process known as Backpropagation. While neural networks are often described through biological metaphors, their true power stems from multi-variable calculus and iterative refinement.

Morning~ On the Continuity of Rotation Representations in Neural Networks (arXiv:1812.07035): ... a topology theorem that deprecated quaternion regression with discontinuities. The recipe: SO(3) ≅ RP³ embeds in R⁵. Surject R⁶ onto SO(3) with Gram-Schmidt process. Simple proof to tell you 4-DoF isn't enough to rotate in 3D space.