728x90 반응형 스터디/인공지능, 딥러닝, 머신러닝56 [CS224W] 2. Feature Engineering for ML in Graphs ML tasks reviewsNode-level prediction → Node featuresLink-level prediction → Link featuresGraph-level prediction → Graph featuresNode-level prediction$G = (V, E)$가 주어질 때, 함수 $f: V \to \mathbb{R}$를 학습한다.Goal: 네트워크의 노드의 구조와 위치를 특성화한다.Node FeaturesNode degreeNode centralityClustering coefficientGraphletsNode Centralitynode degree는 importance를 포착하지 않는다. node centrality $c_v$는 node importance를 지닌다.no.. 2023. 1. 23. [CS224W] 1. Introduction 출처: http://web.stanford.edu/class/cs224w/index.html CS224W: Machine Learningn with Graphs Complex domains (Social networks, internet, Knowledge Graphs, 3D shapes, etc) have a rich relational sturcture, whic can be represented as a relational graph.Modern deep learning toolbox is designed for simple sequences and grids.Not everything can be represented as a sequence or a grid. NN을 이용하여 node의 이웃 .. 2023. 1. 23. 이전 1 ··· 7 8 9 10 다음 728x90 반응형
