Optimization algorithm using matlab

A video tutorial on Firefly Optimization Algorithm and its implementation in MATLAB from scratch

Created byAmir Zaji (Gradient Mentor)

Last updated 9/2018

English

What you'll learn

After watching this video tutorial, you know what is Firefly Optimization Algorithm.
You can use in to solve your own optimization problems.
You can use this optimization algorithm to write your ISI paper.
You know how to write other optimization algorithms in MATLAB.
You can write Firefly Optimization Algorithm in any other programming language that you have experience in.

Course content

3 sections • 17 lectures • 1h 5m total length

What is an optimization problem4:03
Explore optimization problems through regression modeling and neural networks, using weight, height, and walking time to predict blood pressure. Apply MATLAB optimization to minimize error and tune coefficients and weights.
Choosing a function to optimize1:49
Select benchmark optimization problems to study the firefly optimization algorithm, using the sphere function and Rozum Brok function, and implement selected functions in Matlab for the next lecture.
Writing the functions in MATLAB5:17
Write MATLAB functions for the sphere cost and a second optimization function, using preallocation and a for loop, then save as scripts for the Firefly algorithm.

What is Firefly algorithm about?2:15
Firefly algorithm models optimization by light-based attraction, where fireflies communicate with intensity tied to distance via the inverse square law, with linear or surface light source assumptions.
Overview of the Firefly Algorithm procedure1:46
Explore the firefly algorithm for optimization, where fireflies are attracted to brighter peers based on distance and light intensity, iterating with MATLAB to find the best solutions.

Importing the optimization function parameters5:27
Define the optimization problem in Matlab and import the function parameters into the Firefly optimization code, organizing user inputs and variable ranges in a single structure.
Defining the fireflies2:29
Define each firefly as a four-dimensional solution within the -20 to 20 range, compute its cost from its position, and prepare to initialize fireflies for the main optimization loop.
Initializing the fireflies in MATLAB8:41
Initialize the firefly algorithm in Matlab by creating a population structure, setting the number of fireflies, and randomly assigning positions within a defined range; compute costs with the function.
Saving the best Firefly information2:29
Apply a simple matlab firefly rule: after generating positions, save the lowest-cost firefly as the best. Initialize the best cost to infinity.
Formulation of attraction2:47
Present the attractiveness formulation for fireflies, combining inverse square law with gamma absorption. Use this final equation to drive the matlab firefly optimization.
Final attraction equation3:17
Derive the final firefly attraction equation to compute the new position of firefly x from firefly y using attractiveness, distance, gamma, and a random movement epsilon, then implement in Matlab.
Importing the attraction equation in Matlab9:05
Implement the firefly attraction equation in Matlab by defining coefficients, computing distances with norm, updating positions, and preallocating new fireflies for efficiency.
Handling the out of range fireflies2:18
We define how fireflies interact and clamp new positions to a specified range in Matlab, ensuring all fireflies stay within [-20, 24] after movement.
Understanding the Iteration6:53
Explore how the MATLAB firefly optimization iterates to update positions, compare costs, and keep the best firefly through preallocated loops and sorting.
Understanding the damping1:49
Learn how damping the randomization parameter in the firefly optimization algorithm implemented in MATLAB reduces exploration over iterations to boost performance.
Displaying the results4:39
Display the results of the firefly optimization algorithm by plotting the best firefly cost per iteration. Pre-allocate a semilog plot and show convergence toward zero with a sphere example.
MATLAB codes of firefly optimization algorithm0:03

Requirements

Basic knowledge of programming

Description

I’m very glad to have opportunity to teach you one of the most popular and powerful optimization algorithms in this course.

If you search FireFly optimization algorithm in google scholar, it could be seen that there are many vast range of papers has been published by implementing this optimization algorithm in different fields of science.

In this course, after presenting the mathematical concept of each part of the considered optimization algorithm, I write its code immediately in matlab.

All of the written codes are available, however, I strongly suggest to write the codes with me. Notice that, if you don’t have matlab or you know another programming language, don’t worry at all. You can simply write the codes in your own programming language because the behind concepts about all of the written codes are presented completely.

We have a lot to cover in this course, so, let’s start it.

Who this course is for:

Students who are working on optimization algorithms: This course increase their knowledge about one of the most powerful optimization algorithms.
Researchers who want to publish ISI papers in this field: After this course, they can simply apply this optimization algorithm in their own problem.
Anyone who is interested about data mining, Artificial intelligence, and optimization algorithms.

Optimization algorithm using matlab

What you'll learn

Explore related topics

Course content

Basic concepts of an optimization problem3 lectures • 11min

Basic concepts of FireFly Optimization Algorithm2 lectures • 4min

Formulating and coding FireFly Optimization Algorithm12 lectures • 50min

Requirements

Description

Who this course is for: