【Neural-Symbolic】方向的研究工作核心思想一览 Neural-Symbolic可以认为是将人工智能中原本对立的连接主义和符号主义结合的一个新兴研究方向(实际上最早相关工作可追至1978年),先对其相关研究工作典型的十篇论文做出总结。 Neural-Symbolic的本质是什么 Neural-Symbolic意指神经符号主义,但是更本质
论文标题:Neural-Symbolic Computing: An Effective Methodology for Principled Integration of Machine Learning and Reasoning Abstract 有原则性的知识表示和推理机制与基于深度学习的系统相结合,提供合理的、可解释的模型。神经符号计算的目的是,整合两个最基本的认知能力:从环境中学习的能力和从已学知识中推理的能力。
Kuhnberger, Towards integrated neural-symbolic systems for human-level AI: two research programs helping to bridge the gaps, Biol. Inspired Cogn. Architectures 14 (2015) 97-110.Besold T R, Kuhnberger K U. Towards integrated neural-symbolic systems for human-level AI: Two research programs ...
Neural-symbolic reasoning 1. Introduction SYMBOLISM and connectionism are two main paradigms ofArtificial Intelligence. Symbolism assumes the basic units which compose the human intelligence are symbols, and the cognitive process is a series of explicit inferences upon symbolic representations (Minsky and P...
这里就引出来我们第一个想讨论的课题:神经符号学(Neural-Symbolical AI)。 神经符号学-没有定义的算法 回想一下,现在基于学习方式的机器人算法通常具有以下几个问题: 1. 泛化性:基于数据驱动的算法,包括强化学习模型,其效果在很大程度上仍局限于其训练环境,显示出了有限的泛化能力。造成这一限制的原因之一是它们...
Paper reading (三十五):Neural-symbolic machine learning for retrosynthesis and reaction prediction,程序员大本营,技术文章内容聚合第一站。
标题:可解释的视觉问题回答的概率神经符号模型 来源:ICML 2019https://proceedings.mlr.press/v97/vedantam19a.html 代码:https://github.com/kdexd/probnmn-clevr 一、问题提出 neural-symbolic
Neural Symbolic and Probabilistic Logic research explores integrating logic with neural networks. Key areas include surveys, logic-enhanced neural networks, mod...
neurallambda/neurallambda Star246 Code Issues Pull requests Discussions Reasoning Computers. Lambda Calculus, Fully Differentiable. Also Neural Stacks, Queues, Arrays, Lists, Trees, and Latches. machine-learningagirnnreasoningneural-symbolic-reasoningneural-symbolic ...
"GNNs endowed with attention mechanisms are a promising direction of research towards the provision of rich reasoning and learning in type 6 neural- symbolic systems", 还有好多问题:1. N现在比较明确了2. S是是符号、嵌入表示、图?3. N如何和S结合?4. 什么时候唤起S?5. N和S在具体不同任务中的形...