ACO & AIS for Solving Combinatorial Optimization Problems

CSE 462 Algorithm Engineering CSE, BUET

FB

ACO & AIS

ACO: Ant Colony Optimization
Inspired by the natural behavior of antz

AIS: Artificial Immune System

Inspired by biological immunity system

FB

Ant Colony Optimization

Optimization Technique Proposed by Marco Dorigo in the early 90 Discrete optimization problems difficult to solve Modeling ant colony behavior Often applied to TSP (Travelling Salesman Problem): shortest path between n nodes

FB

Ant Colony Optimization

The pheromone concentration on trail B will increase at a higher rate than on A, and soon the ants on route A will choose to follow route B Since most ants will no longer travel on route A, and since the pheromone is volatile, trail A will start evaporating Only the shortest route will remain!
FB

Since the route B is shorter, the ants on this path will complete the travel more times and thereby lay more pheromone over it.

TSP: Travelling Salesman Problem

Given N cities, and a distance function d between cities, find a tour that: 1. Goes through every city once and only once 2. Minimizes the total distance.

FB

ACO for TSP

Ants (blind) navigate from nest to food source

Shortest path is discovered via pheromone trails

each ant moves at random pheromone is deposited on path more pheromone on path increases probability of path being followed

FB

ACO for TSP

Starting node selected at random

Path selected at random

based on amount of trail present on possible paths from starting node

higher probability for paths with more trail

Ant reaches next node, selects next path Continues until reaches starting node Finished tour is a solution
FB

ACO for TSP

A completed tour is analyzed for optimality Trail amount adjusted to favor better solutions
better solutions receive more trail worse solutions receive less trail higher probability of ant selecting path that is part of a better-performing tour

New cycle is performed Repeated until most ants select the same tour on every cycle (convergence to solution)

FB

ACO for TSP

Algorithm in Pseudo code:
Initialize Trail Do While (Stopping Criteria Not Satisfied) Cycle Loop Do Until (Each Ant Completes a Tour) Tour Loop
Local Trail Update

End Do Analyze Tours Global Trail Update End Do

FB

ACO
Can be used for both Static and Dynamic Combinatorial optimization problems Convergence is guaranteed, although the speed is unknown
Value Solution

FB

ACO Parameters
The parameters considered here are those that affect directly or indirectly the computation of the probability for path transition:
: the relative importance of the trail, >0 : the relative importance of the visibility, >0 : trail persistence, Q : a constant related to the quantity of trail laid by ants as trail evaporation

FB

Visibility in ACO
A local heuristic function related to the problem domain
A function based on d in case of TSP

Typically local heuristic functions are similar to objective functions in greedy algorithms.

FB

ACO Stopping Conditions

Stagnation Max Iterations

FB

ACO for TSP

FB

ACO for MKP and MMKP

MKP: Multidimensional Knapsack Problem
Each object requires multiple resources For each resource there is a constraint

MMKP: Multichoise Multidimensional Knapsack Problem

Each object requires multiple resources For each resource there is a constraint

Objects are arranged in groups Exactly one object from each group
FB

ACO and MKP

3 different approaches depending on pheromone trail location
On objects On pair of objects On set of objects

Heuristic functions depending on the ratio of profit and resource consumption rate of an object
FB

ACO and MMKP

Pheromone is laid on objects Heuristic function

A random local search to improve solution quality

FB

ACO and MMKP

FB

ACO and MMKP

Approaches under study
Group ordering Guided local search Hybrid approaches (GA, Bee Colony, AIS)

FB

FB

AIS: Artificial Immune System

Exploit the immune system's characteristics of learning and memory to solve a problem. List of some Algorithms and Models
Bone Marrow Models Negative Selection Algorithms Clonal Selection Algorithm Somatic Hypermutation Immune Network Models Dendritic Cell Algorithms
FB

AIS: Artificial Immune System

CLONALG uses two populations:
Antigens Antibodies.

When used for optimization, the main idea of CLONALG is to

reproduce individuals with a high affinity apply mutation (or blind variation) select the improved maturated progenies produced.
FB

AIS: Artificial Immune System

Immune system has some remarkable properties Pattern recognition Learning Memory So, is it useful?

FB

AIS for KPs

FB

FB

FB

AIS for KPs

A lot of scope for improvement
!! CLONALG !! Local search procedures from ANTZ can be used here Group ordering can be incorporated

FB

Thank you !

FB