File Exchange

image thumbnail

Teaching Learning Based Optimization

version 1.2.0.0 (2.21 KB) by SKS Labs
Correct implementation of TLBO without the removal of duplicates

22 Downloads

Updated 25 Jun 2018

View License

Teaching Learning Based Optimization is a single objective optimization technique for unconstrained problems.
In TLBO, as proposed in literature, a student has to complete both the teacher and the learner phase. This is followed by the subsequent student undergoing the teacher and the student phase. However in most implementations [1,2], all the members are required to initially complete the teacher phase and subsequently all the members undergo the student phase. This issue has been reported in literature.
http://www.sciencedirect.com/science/article/pii/S0020025512003532

The discrepancy arises as the authors have not depicted the population loop in their flowchart/pseudo code. There is also a discrepancy in the code provided by the authors of TLBO and also the examples given in the articles.

Moreover, the TLBO as proposed in the literature requires the identification and replacement of duplicate solutions in every iteration. However the occurrence of duplicate solutions is limited in problems with reasonable number of decision variables. Thus the computational load of identification of duplicate solutions can be eliminated and also the limited functional evaluations can be effectively utilized. To the best of our knowledge, no other single objective evolutionary technique requires the identification and removal of duplicate solutions. Hence this implementation does not identify and remove the duplicates from the population in every iteration.

Note:
i) If N is the size of the population, exactly 2N functional evaluations are required in a complete iteration of TLBO. If there are I iterations, then the total number of functional evaluations will be 2NI + N. This is because 2NI evaluations are required for all the iteration and the N population members are to be evaluated before the start of the iteration.

(ii) TLBO has two user defined parameters, viz., (i) Population (or class) Size and (ii) the number of iterations (or an alternate termination criterion).

(iii) TLBO is monotonically convergent.

(1) https://in.mathworks.com/matlabcentral/fileexchange/51202-teaching-learning-based-optimization-for-truss-optimization
(2) https://in.mathworks.com/matlabcentral/fileexchange/52863-teaching-learning-based-optimization--tlbo-

Cite As

SKS Labs (2020). Teaching Learning Based Optimization (https://www.mathworks.com/matlabcentral/fileexchange/65628-teaching-learning-based-optimization), MATLAB Central File Exchange. Retrieved .

Comments and Ratings (5)

rajesh devineni

Obaid Ullah

Sonam Gupta

Maheen Khan

Raphael Zimmermann

Updates

1.2.0.0

Update comment and formatting.

1.1.0.0

Updated file name.

MATLAB Release Compatibility
Created with R2017b
Compatible with any release
Platform Compatibility
Windows macOS Linux