MLG笔记:(一) Motivation for Graph ML

Why Graph? Graph體現了entities(nodes)之間的聯繫,這些聯繫構成一個network。許多的數據都可以自然的用graph來建模表示,例如計算機網路、疾病傳染路徑、社交網路等。即graph有廣泛的應用空間。 課程討論的主要問題: How do we take advantag...

論文:Machine Learning on Graphs: A Model and Comprehensive Taxonomy

貢獻 幾位研究者提出了GRAPHEDM(Graph Encoder Decoder Model)模型,將semi-supervised learning(e.g. GraphSage,GCN,GAT)與unsupervised learning of graph representations(e.g....

LA筆記:(二十) Eigenvalues, Eigenvectors and Diagonalization

Def of Eigenvalue and Eigenvector: Let be a n-by-n matrix. A scalar is called an eigenvalue of if there exist a nonzero vector such that The vec...

LA筆記:(十九) Matrix Transformation from R(dimension-n) to R(dimension-m)

Def of Linear transformation: Let and be vector spaces(over ). A function is called a linear transformation from into if for all and such that: &...

LA筆記:(十八) Rank, Nullity and Fundamental Matrix Space

Theorem: The and have the same dimension.(In spite of the size of ) Proof: Let be the RREF of Number of leading1's variables Def of ...

LA筆記:(十七) Row Space, Column Space and Null space

Def of Row vector and column vector: For any m-by-n matrix , the vectors ,,, in that are formed from the rows of are called the row vectors of , and ...

LA筆記:(十六) Dimension

Dimension(向量空間維度) Theorem(基底向量個數恆定): Let be a finite-dimensional vector space and let be a basis for , then <a>If a set in that has more tha...

LA筆記:(十五) Coordinates and Basis

basis(基底) Def of basis: If is any vector space and is a subset of , then is called a basis for if the following two conditions hold: <a> i...

LA筆記:(十四) Linear Independence

線性獨立 Def: If is a nonempty set of a vector space , then the vector equation has at least one solution, the trivial solution. If this is the only sol...

LA筆記:(十三) Subspace

向量空間子空間定義 A subset of a vector space $V is called a subspace of is is itself a vector space under the addition and scalar multiplication defined on ....