Artificial intelligence - TSP
This document discusses using genetic algorithms to help a salesman solve the traveling salesman problem of planning the shortest route to visit all cities. It introduces the traveling salesman problem and explains how genetic algorithms could help by simulating evolution to iteratively find better routes. Specifically, it would initialize a population of routes, select the fittest for reproduction, crossover routes to create new offspring, occasionally mutate routes, evaluate all routes, and repeat the process until a termination condition is met, such as reaching a set number of iterations. This evolutionary approach allows finding acceptable solutions within time constraints compared to traditional brute force methods for this complex optimization problem.
Artificial intelligence - TSP
- 1. Artificial IntelligenceGenetic Algorithms and Travel Salesman ProblemGiáo viên hướng dẫn: Trần Cao TrưởngSinh viên thực hiện: Nguyễn Đức Hiển Lê Hữu Sơn Tùng
- 2. Artificial IntelligencePowerPoint has new layouts that give you more ways to present your words, images and media. Fun fact about salesmen todayGenetic Algorithms(GAs) and Travel Salesman Problem(TSP)
- 3. 1. TSP – Overview(p.1)Hello eveyone ! I’m a HANDSOME salesman.
- 4. 1. TSP – Overview(p.2)As a salesman, I have to travel through many cities for business. But I just need to travel EACH city ONCE.City 1City 3City 2City n-1City n-2MAPCity n
- 5. 1. TSP – Overview(p.3) I have some problems. I don’t have much time and money.
- 6. 1. TSP – Overview(p.4) Can you help me to find the way with minimum cost ?
- 7. 1. TSP – Traditional Solutions(p.1)BRUTE- FORCE
- 8. 2.GAs – Overview (p.1) GAs simulate the evolution of one population (for example human population) : based on assessing and selecting the fittest solutions in the problem population.So GAs has some similar components as an evolution in the real world: + POPULATION+ INDIVIDUALS+ CHROMOSOMES+ OFFSPRING+ CROSSOVER+ MUTATION + FITNESS
- 9. 2.GAs – Overview (p.2)GAs
- 10. 2.GAs – Why use GAs to solve TSP ? (p.1)Reasons to use GAs to solve TSP includes the followings: Because the map of cities in TSP is a complete graph so searching space in TSP is very huge.It’s possbile for GAs to find out an acceptable solution due to given conditions and constraints(ex : time constraint, etc.).
- 11. 3.Using GAs to solve TSP? (p.1)A generation in TSP includes the following components:
- 12. 3.Using GAs to solve TSP? (p.2)GA Structure:GA( ){ t=0;// epoch Initialize(t); Evaluate(t); While (not termination condition) do { t= t+1; Select P(t) from P(t-1); Alter P(t); Evaluate P(t); }}Step 0: InitializationStep 1: SelectionStep 2: CrossoverStep 3: MutationStep 4: EvaluationStep 5: Termination TestStep 0: InitializationStep 6: End
- 13. 3.GAs – Why use GAs to solve TSP ? (p.3)Processes : Initialize the population : Randomize cities with following attribute: + Name or IDs : to distinguish cities.+ Link between 2 cities.+ Each link has a weighted- number represented as the cost to travel from one city to another.
- 14. 3.GAs – Why use GAs to solve TSP ? (p.4)2. Loop(Epoch): + Choosing 2 parents from from a random group of cities.+ Crossover -> Offsprings (2)+ Mutation+ Evaluation + Use fitness function to remove some weaker individuals to keep the population size constant.
- 15. 3.GAs – Why use GAs to solve TSP ? (p.5)3. Terminate Process of Finding Solution:Based on some termination condition (ex : Limitation of Iterations) , terminate the process.Fun fact about Salesmen todayYears ago, this is the way old men used to solve TSP.
- 16. Fun fact about Salesmen todayToday, young men think different !!!
- 17. Thank you for your attention !!!