Algorithmic Primitives for Network AnalysisThrough the Lens of the Laplacian Paradigm网络分析的算法原语： 通过拉普拉斯范式的透镜.pdfVIP

Algorithmic Primitives for Network Analysis: Through the Lens of the Laplacian Paradigm Shang-Hua Teng Computer Science, Viterbi School of Engineering USC Massive Data and Massive Graphs • 500 billions web pages • unbounded amount of web logs • billions of variables • billions of transistors. Happy Asymptotic World for Theoreticians Efficient Algorithm Challenges Nearly-Linear-Time Algorithms Quadratic time algorithms could be too slow!!!! Algorithmic Paradigms • Greedy often nearly linear (limited applications) • Dynamic Programming usually not linear (even when applicable) • Divide-and-Conquer sometimes nearly linear • Mathematical Programming rarely linear • Branch-and-Bound hardly linear • Multilevel Methods mostly linear (lack of proofs) • Local Search and Simulated Annealing can be nearly linear Examples: Nearly-Linear-Time Graph Algorithms Breadth-First Search O(|V|+|E|) Depth-First Search Shortest Path Tree Minimum Spanning Tree Planarity Testing Bi-connected components Topological sorting Sparse matrix vector product Algorithm Design is like Building a Software Library Once a nearly-linear time (sub-linear time) algorithm is developed, it can be used as a primitive or a subroutine in the design of other nearly-linear-time (sub-linear time) algorithms. Laplacian Primitive Solve A x = b , where A is a weighted Laplacian matrix Laplacian Primitive Solve A x = b , where A is a weighted Laplacian matrix A is Laplacian matrix: symmetric non-posi