The Steepest Ascent Hill Climbing Algorithm

Slides:



Advertisements
Similar presentations
Lecture 7. Network Flows We consider a network with directed edges. Every edge has a capacity. If there is an edge from i to j, there is an edge from.
Advertisements

15.082J & 6.855J & ESD.78J October 14, 2010 Maximum Flows 2.
MAXIMUM FLOW Max-Flow Min-Cut Theorem (Ford Fukerson’s Algorithm)
1 Augmenting Path Algorithm s t G: Flow value = 0 0 flow capacity.
Chapter 10: Iterative Improvement The Maximum Flow Problem The Design and Analysis of Algorithms.
Lectures on Network Flows
1 COMMONWEALTH OF AUSTRALIA Copyright Regulations 1969 WARNING This material has been reproduced and communicated to you by or on behalf.
1 Augmenting Path Algorithm s t G: Flow value = 0 0 flow capacity.
Maximum Flows Lecture 4: Jan 19. Network transmission Given a directed graph G A source node s A sink node t Goal: To send as much information from s.
Flows sourcesink s t Flows sourcesink edge-weights = capacities.
DAST 2005 Week 4 – Some Helpful Material Randomized Quick Sort & Lower bound & General remarks…
Minimal Spanning Trees What is a minimal spanning tree (MST) and how to find one.
Theory of Computing Lecture 10 MAS 714 Hartmut Klauck.
Single Source Shortest-Path: The General Case (with negative edges) Bellman-Ford algorithm. Iteratively relax all edges |V|-1 times Running time? O(VE).
1 COMMONWEALTH OF AUSTRALIA Copyright Regulations 1969 WARNING This material has been reproduced and communicated to you by or on behalf.
Prim's Algorithm This algorithm starts with one node. It then, one by one, adds a node that is unconnected to the new graph to the new graph, each time.
Algorithm Design and Analysis (ADA)
1 EE5900 Advanced Embedded System For Smart Infrastructure Static Scheduling.
NETWORK FLOWS Shruti Aggrawal Preeti Palkar. Requirements 1.Implement the Ford-Fulkerson algorithm for computing network flow in bipartite graphs. 2.For.
Theory of Computing Lecture 12 MAS 714 Hartmut Klauck.
A directed graph G consists of a set V of vertices and a set E of arcs where each arc in E is associated with an ordered pair of vertices from V. V={0,
Maximum Flow Problem Definitions and notations The Ford-Fulkerson method.
1 Network Flow & Linear Programming Jeff Edmonds York University COSC 3101 Lecture 5 Optimization Problems Network Flow Def n Network Flow Def n Simple.
Network Flow & Linear Programming Jeff Edmonds York University COSC 6111 Lecture 3 Network Flow Linear Programming.
Minimum Cost Flow Algorithms and Networks. Algorithms and Networks: Minimum Cost Flow2 This lecture The minimum cost flow problem: statement and applications.
TU/e Algorithms (2IL15) – Lecture 8 1 MAXIMUM FLOW (part II)
Maximum Flow c v 3/3 4/6 1/1 4/7 t s 3/3 w 1/9 3/5 1/1 3/5 u z 2/2
Max-flow, Min-cut Network flow.
Data Structures and Algorithms (AT70. 02) Comp. Sc. and Inf. Mgmt
BIPARTITE GRAPHS AND ITS APPLICATIONS
Randomized Algorithm (Lecture 2: Randomized Min_Cut)
Finding a Path With Largest Smallest Edge
Chapter 7: Greedy Algorithms
CS4234 Optimiz(s)ation Algorithms
Algorithm Design and Analysis
Lectures on Network Flows
Network Flow & Linear Programming
Algorithms and Networks
CSCI 3160 Design and Analysis of Algorithms Tutorial 8
I206: Lecture 15: Graphs Marti Hearst Spring 2012.
Topological Sort (topological order)
More Graph Algorithms.
Graph Algorithm.
Max-flow, Min-cut Network flow.
Maximum Flow Problem flow capacity Actual flow  capacity
Lecture 16 Maximum Matching
Network Flow 2016/04/12.
Instructor: Shengyu Zhang
Edmonds-Karp Algorithm
Maximum Flow c v 3/3 4/6 1/1 4/7 t s 3/3 w 1/9 3/5 1/1 3/5 u z 2/2
Lecture 19-Problem Solving 4 Incremental Method
Shortest Path Algorithms
Flow Networks Topics Flow Networks Residual networks
Augmenting Path Algorithm
Route Inspection Which of these can be drawn without taking your pencil off the paper and without going over the same line twice? If we introduce a vertex.
Flow Networks General Characteristics Applications
Flow Networks and Bipartite Matching
Complexity of Ford-Fulkerson
Algorithms (2IL15) – Lecture 7
EE5900 Advanced Embedded System For Smart Infrastructure
Algorithms and Networks
Maximum Flow c v 3/3 4/6 1/1 4/7 t s 3/3 w 1/9 3/5 1/1 3/5 u z 2/2
Lecture 21 Network Flow, Part 1
MAXIMUM flow by Eric Wengert.
Augmenting Path Algorithm
Maximum Flow Neil Tang 4/8/2008
Lecture 12 Shortest Path.
Maximum Bipartite Matching
Advanced Graph Homer Lee 2013/10/31.
Maximum Flow Problems in 2005.
Presentation transcript:

The Steepest Ascent Hill Climbing Algorithm Muna Shabaneh

Instead of winding back and forth in hill climbing algorithm Steepest Ascent: you always take the step that increases your height the most

Network Flow 45 a 25 3 b s 3 t 10 40 5 c 4 We will try to find a path in the augmenting graph that would give us the Biggest Smallest Edge between all other paths from s to t.

Steepest Ascent Hence the amount the flow increases is the smallest augmentation capacity of any edge in this path. The Steepest Ascent Algorithm will augment this best path in each iteration. Our algorithm should be able to find such a path  and it will find the maximum flow within a instead of iterations. ( that is what we will prove)

Finding the augmenting path with the Biggest Smallest Edge: The Input: 45 G: a 25 3 b s 3 t 10 40 5 c 4 The Output: The best path from s to t whose smallest weighted edge is as big as possible.

Easier Problem Input : 1) A path from s o t whose smallest weighted edge <u,v> ≥ 2) No path : Is an input weight, we can give it any value In our augmented graph any edge whose weight less than will be deleted from the augmented graph.

First we will sort the edges from biggest to smallest, after that we will choose our initial weight to be the first weight value and we will draw an edge for every weight that is ≥ . weight Edges 45 a t 1 40 s c 2 25 s a 3 10 b a 4 5 c b c t 6 a b 7 a c = = 45 = = 40 = = 25

The augmented graph a b s t c No path No path 45 No path a 25 b s t 40 c No path A path with the Biggest Smallest Edge < s, a, t > = F <25>

Running Time The algorithm for finding the path with the largest smallest edge runs Depth-First search for each weight w. We reuse the previous work so no node is marked reached more than once and no edge is traversed more than once. So we need only O(m) time for this process m : number of edges. And we need O( log m ) for sorting the edges. So we need O(m log m) for finding the biggest smallest edge.

Running Time for Steepest Ascent F = 0 initially and then it may increase up to O( m. ) m : number of edges l : bit capacities. : The maximum flow for the network. rate (F) : The current flow : the remaining amount the flow should increase after iteration. So the measure of progress = - rate (F) : The maximum flow for the network. rate (F) : The current flow : the remaining amount the flow should increase after iteration. So the measure of progress = - rate (F) OOPS WE HAVE A PROBLEM ?? ???????????????????

Choosing a Cut C = <U,V> The sum of the augmentation capacities of the edges across any cut C=<U,V> acts as an upper bound to the total flow possible. =  The smallest augmentation capacity in the path whose smallest augmentation capacity is the largest. from and we know that the edge which was deleted in and now in has augmentation capacity at most The sum of the augmentation capacities in the cut C=<U,V> is at most m . Number of Iterations : O(ml + m log m) Finally the running time O(ml+m log m) x O(m log m) = O (l log m+ m)

We Did It !!!!!!