Ant Colony Optimization (#1237)

core-java/src/main/java/com/baeldung/algorithms/RunAlgorithm.java

@@ -2,7 +2,8 @@ package com.baeldung.algorithms;
 
 import java.util.Scanner;
 
-import com.baeldung.algorithms.annealing.SimulatedAnnealing;
+import com.baeldung.algorithms.ga.annealing.SimulatedAnnealing;
+import com.baeldung.algorithms.ga.ant_colony.AntColonyOptimization;
 import com.baeldung.algorithms.ga.binary.SimpleGeneticAlgorithm;
 import com.baeldung.algorithms.slope_one.SlopeOne;
 
@@ -14,6 +15,7 @@ public class RunAlgorithm {
 		System.out.println("1 - Simulated Annealing");
 		System.out.println("2 - Slope One");
 		System.out.println("3 - Simple Genetic Algorithm");
+		System.out.println("4 - Ant Colony");
 		int decision = in.nextInt();
 		switch (decision) {
 		case 1:
@@ -27,6 +29,10 @@ public class RunAlgorithm {
 			SimpleGeneticAlgorithm ga = new SimpleGeneticAlgorithm();
 			ga.runAlgorithm(50, "1011000100000100010000100000100111001000000100000100000000001111");
 			break;
+		case 4:
+			AntColonyOptimization antColony = new AntColonyOptimization(21);
+			antColony.startAntOptimization();
+			break;
 		default:
 			System.out.println("Unknown option");
 			break;

core-java/src/main/java/com/baeldung/algorithms/ga/annealing/City.java

@@ -1,4 +1,4 @@
-package com.baeldung.algorithms.annealing;
+package com.baeldung.algorithms.ga.annealing;
 
 import lombok.Data;
 

core-java/src/main/java/com/baeldung/algorithms/ga/annealing/SimulatedAnnealing.java

@@ -1,4 +1,4 @@
-package com.baeldung.algorithms.annealing;
+package com.baeldung.algorithms.ga.annealing;
 
 public class SimulatedAnnealing {
 

core-java/src/main/java/com/baeldung/algorithms/ga/annealing/Travel.java

@@ -1,4 +1,4 @@
-package com.baeldung.algorithms.annealing;
+package com.baeldung.algorithms.ga.annealing;
 
 import java.util.ArrayList;
 import java.util.Collections;

core-java/src/main/java/com/baeldung/algorithms/ga/ant_colony/Ant.java

@@ -0,0 +1,37 @@
+package com.baeldung.algorithms.ga.ant_colony;
+
+public class Ant {
+
+	protected int trailSize;
+	protected int trail[];
+	protected boolean visited[];
+
+	public Ant(int tourSize) {
+		this.trailSize = tourSize;
+		this.trail = new int[tourSize];
+		this.visited = new boolean[tourSize];
+	}
+
+	protected void visitCity(int currentIndex, int city) {
+		trail[currentIndex + 1] = city;
+		visited[city] = true;
+	}
+
+	protected boolean visited(int i) {
+		return visited[i];
+	}
+
+	protected double trailLength(double graph[][]) {
+		double length = graph[trail[trailSize - 1]][trail[0]];
+		for (int i = 0; i < trailSize - 1; i++) {
+			length += graph[trail[i]][trail[i + 1]];
+		}
+		return length;
+	}
+
+	protected void clear() {
+		for (int i = 0; i < trailSize; i++)
+			visited[i] = false;
+	}
+
+}

core-java/src/main/java/com/baeldung/algorithms/ga/ant_colony/AntColonyOptimization.java

@@ -0,0 +1,212 @@
+package com.baeldung.algorithms.ga.ant_colony;
+
+import java.util.Arrays;
+import java.util.Random;
+
+public class AntColonyOptimization {
+
+	private double c = 1.0;
+	private double alpha = 1;
+	private double beta = 5;
+	private double evaporation = 0.5;
+	private double Q = 500;
+	private double antFactor = 0.8;
+	private double randomFactor = 0.01;
+
+	private int maxIterations = 1000;
+
+	public int numberOfCities;
+	public int numberOfAnts;
+	private double graph[][];
+	private double trails[][];
+	private Ant ants[];
+	private Random random = new Random();
+	private double probabilities[];
+
+	private int currentIndex;
+
+	public int[] bestTourOrder;
+	public double bestTourLength;
+
+	public AntColonyOptimization(int noOfCities) {
+		graph = generateRandomMatrix(noOfCities);
+		numberOfCities = graph.length;
+		numberOfAnts = (int) (numberOfCities * antFactor);
+
+		trails = new double[numberOfCities][numberOfCities];
+		probabilities = new double[numberOfCities];
+		ants = new Ant[numberOfAnts];
+		for (int j = 0; j < numberOfAnts; j++) {
+			ants[j] = new Ant(numberOfCities);
+		}
+	}
+
+	/**
+	 * Generate initial solution
+	 * @param n
+	 * @return
+	 */
+	public double[][] generateRandomMatrix(int n) {
+		double[][] randomMatrix = new double[n][n];
+		random.setSeed(System.currentTimeMillis());
+		for (int i = 0; i < n; i++) {
+			for (int j = 0; j < n; j++) {
+				Integer r = random.nextInt(100) + 1;
+				randomMatrix[i][j] = Math.abs(r);
+			}
+		}
+		return randomMatrix;
+	}
+	
+	/**
+	 * Perform ant optimization
+	 * @return
+	 */
+	public int[] startAntOptimization() {
+		int[] finalResult = null;
+		for (int i = 1; i <= 3; i++) {
+			System.out.println("Attempt #" + i);
+			finalResult = solve();
+		}
+		return finalResult;
+	}
+	
+	/**
+	 * Use this method to run the main logic
+	 * @return
+	 */
+	private int[] solve() {
+		setupAnts();
+		clearTrails();
+		int iteration = 0;
+		while (iteration < maxIterations) {
+			moveAnts();
+			updateTrails();
+			updateBest();
+			iteration++;
+		}
+		System.out.println("Best tour length: " + (bestTourLength - numberOfCities));
+		System.out.println("Best tour order: " + Arrays.toString(bestTourOrder));
+		return bestTourOrder.clone();
+	}
+	
+	/**
+	 * Prepare ants for the simulation
+	 */
+	private void setupAnts() {
+		currentIndex = -1;
+		for (int i = 0; i < numberOfAnts; i++) {
+			ants[i].clear();
+			ants[i].visitCity(currentIndex, random.nextInt(numberOfCities));
+		}
+		currentIndex++;
+	}
+	
+	/**
+	 * At each iteration, move ants
+	 */
+	private void moveAnts() {
+		while (currentIndex < numberOfCities - 1) {
+			for (Ant a : ants)
+				a.visitCity(currentIndex, selectNextCity(a));
+			currentIndex++;
+		}
+	}
+	
+	/**
+	 * Select next city for each ant
+	 * @param ant
+	 * @return
+	 */
+	private int selectNextCity(Ant ant) {
+		if (random.nextDouble() < randomFactor) {
+			int t = random.nextInt(numberOfCities - currentIndex);
+			int j = -1;
+			for (int i = 0; i < numberOfCities; i++) {
+				if (!ant.visited(i)) {
+					j++;
+				}
+				if (j == t) {
+					return i;
+				}
+			}
+		}
+		calculateProbabilities(ant);
+		double r = random.nextDouble();
+		double total = 0;
+		for (int i = 0; i < numberOfCities; i++) {
+			total += probabilities[i];
+			if (total >= r) {
+				return i;
+			}
+		}
+
+		throw new RuntimeException("There are no other cities");
+	}
+
+	/**
+	 * Calculate the next city picks probabilites
+	 * @param ant
+	 */
+	private void calculateProbabilities(Ant ant) {
+		int i = ant.trail[currentIndex];
+		double pheromone = 0.0;
+		for (int l = 0; l < numberOfCities; l++) {
+			if (!ant.visited(l)) {
+				pheromone += Math.pow(trails[i][l], alpha) * Math.pow(1.0 / graph[i][l], beta);
+			}
+		}
+		for (int j = 0; j < numberOfCities; j++) {
+			if (ant.visited(j)) {
+				probabilities[j] = 0.0;
+			} else {
+				double numerator = Math.pow(trails[i][j], alpha) * Math.pow(1.0 / graph[i][j], beta);
+				probabilities[j] = numerator / pheromone;
+			}
+		}
+	}
+
+	/**
+	 * Update trails that ants used
+	 */
+	private void updateTrails() {
+		for (int i = 0; i < numberOfCities; i++) {
+			for (int j = 0; j < numberOfCities; j++) {
+				trails[i][j] *= evaporation;
+			}
+		}
+		for (Ant a : ants) {
+			double contribution = Q / a.trailLength(graph);
+			for (int i = 0; i < numberOfCities - 1; i++) {
+				trails[a.trail[i]][a.trail[i + 1]] += contribution;
+			}
+			trails[a.trail[numberOfCities - 1]][a.trail[0]] += contribution;
+		}
+	}
+
+	/**
+	 * Update the best solution
+	 */
+	private void updateBest() {
+		if (bestTourOrder == null) {
+			bestTourOrder = ants[0].trail;
+			bestTourLength = ants[0].trailLength(graph);
+		}
+		for (Ant a : ants) {
+			if (a.trailLength(graph) < bestTourLength) {
+				bestTourLength = a.trailLength(graph);
+				bestTourOrder = a.trail.clone();
+			}
+		}
+	}
+
+	/**
+	 * Clear trails after simulation
+	 */
+	private void clearTrails() {
+		for (int i = 0; i < numberOfCities; i++)
+			for (int j = 0; j < numberOfCities; j++)
+				trails[i][j] = c;
+	}
+
+}

core-java/src/test/java/com/baeldung/algorithms/AntColonyOptimizationTest.java

@@ -0,0 +1,22 @@
+package com.baeldung.algorithms;
+
+import org.junit.Assert;
+import org.junit.Test;
+
+import com.baeldung.algorithms.ga.ant_colony.AntColonyOptimization;
+
+public class AntColonyOptimizationTest {
+
+	@Test
+	public void testGenerateRandomMatrix() {
+		AntColonyOptimization antTSP = new AntColonyOptimization(5);
+		Assert.assertNotNull(antTSP.generateRandomMatrix(5));
+	}
+
+	@Test
+	public void testStartAntOptimization() {
+		AntColonyOptimization antTSP = new AntColonyOptimization(5);
+		Assert.assertNotNull(antTSP.startAntOptimization());
+	}
+
+}

core-java/src/test/java/com/baeldung/algorithms/SimulatedAnnealingTest.java

@@ -3,7 +3,7 @@ package com.baeldung.algorithms;
 import org.junit.Assert;
 import org.junit.Test;
 
-import com.baeldung.algorithms.annealing.SimulatedAnnealing;
+import com.baeldung.algorithms.ga.annealing.SimulatedAnnealing;
 
 public class SimulatedAnnealingTest {