Ten Artificial Human Optimization Algorithms

: The term “Artificial Human Optimization” was first coined by the corresponding author of this work in December 2016 when he published a paper titled “Entrepreneur : Artificial Human Optimization” at Transactions on Machine Learning and Artificial Intelligence (TMLAI) Volume 4, No 6 (December 2016). According to that paper published in 2016, Artificial Human Optimization Field is defined as the collection of all those optimization algorithms which were proposed based on Artificial Humans. In real world we (Humans) solve the problems. In the same way Artificial Humans imitate real Humans in the search space and solve the optimization problems. In Particle Swarm Optimization (PSO) the basic entities in the solution space are Artificial Birds where as in Artificial Human Optimization the basic entities in search space are Artificial Humans. Each Artificial Human corresponds to a point in the solution space. Ten Artificial Human Optimization methods titled “Human Bhagavad Gita Particle Swarm Optimization (HBGPSO)”, “Human Poverty Particle Swarm Optimization (HPPSO)”, “Human Dedication Particle Swarm Optimization (HuDePSO)”, “Human Selection Particle Swarm Optimization (HuSePSO)”, “Human Safety Particle Swarm Optimization (HuSaPSO)”, “Human Kindness Particle Swarm Optimization (HKPSO)”, “Human Relaxation Particle Swarm Optimization (HRPSO)”, “Multiple Strategy Human Particle Swarm Optimization (MSHPSO)”, “Human Thinking Particle Swarm Optimization (HTPSO)”, “Human Disease Particle Swarm Optimization (HDPSO)” are applied on various benchmark functions and results obtained are shown in this work.


Introduction
The goal of 'Human Optimization' is to increase the performance of real humans through various methods. But 'Artificial Human Optimization' is a new field which took its birth recently in December 2016 as explained in abstract of this paper. This new filed is a sub-field of Evolutionary Computing which in turn is a sub-field of Computational Intelligence field. Hence 'Human Optimization (Real Human Optimization)' is different from Artificial Human Optimization (AHO).
The following is the review obtained from an expert in 2013 for a work under AHO Field. The review is shown below in double quotes as it is: "The motivation of the paper is interesting. But the paper does not present any evaluation of the proposed algorithm. So we have an idea but we are not able to assess it on the basis of the paper. Next, there seems to be a difference between birds, fishes, ants, bacteria, bees etc. on one side, and human beings on the other side. Birds, fishes, ants, bacteria, bees etc. are more or less the same. People are different. I dare say that taxi drivers are different from politicians, or preschool teachers for example. Some people prefer money or power than love. It is not so difficult to guess which way ants will go but it is not so obvious when we consider people behavior. In my opinion the paper is a very first step to build the algorithm assumed but still lots of work is needed to achieve the goal." The algorithms under Artificial Human Optimization Field (AHO Field) were proposed in literature starting from year 2003. But from the above review it is clear that the expert felt there are no algorithms under Artificial Human Optimization Field as on 2013 and corresponding author's work is the very first step. Experts are very familiar with Genetic Algorithms, Particle Swarm Optimization, Ant Colony Optimization etc but according to corresponding author's observation many experts are unaware of the fact that there are algorithms under AHO Field before 2013. Even corresponding author of this work felt that his work submitted for review in 2013 is the beginning of Artificial Human Optimization Field Algorithms. But this was a mistake and it was corrected in later papers. It is also clear from above review shown in double quotes that imitating Humans and creating Evolutionary Computing algorithms is not as easy as imitating beings Birds, fishes, ants, bacteria, bees etc and creating algorithms under Evolutionary Computation domain.

Particle Swarm Optimization
Particle Swarm Optimization (PSO) was proposed by Kennedy and Eberhart in 1995. PSO is based on Artificial Birds. It has been applied to solve complex optimization problems.
In PSO, first we initialize all particles as shown below. Two variables pbesti and gbest are maintained. pbesti is the best fitness value achieved by i th particle so far and gbest is the best fitness value achieved by all particles so far. Lines 4 to 11 in the below text helps in maintaining particle best and global best. Then the velocity is updated by rule shown in line no. 14. Line 15 updates position of i th particle. Line 19 increments the number of iterations and then the control goes back to line 4. This process of a particle moving towards its local best and also moving towards global best of particles is continued until termination criteria will be reached. for each particle i do

Human Bhagavad Gita Particle Swarm Optimization
Bhagavad Gita is a Hindu sacred text. There are no Hybrid PSO algorithms based on Bhagavad Gita till date. According to Bhagavad Gita "He who is successful is not ideal. He who failed is not ideal. Only he is ideal and revered who irrespective of success or failure stands steadfast in the pursuit of his mission". Human Bhagavad Gita Particle Swarm Optimization (HBGPSO) is designed based on this fact.
The population consists of ideal and non ideal candidates. Based on random number generated and IdealCandidateProbability, the human is classified into either ideal or non ideal candidate. Ideal candidate is not affected by success or failure and he moves in search space without any halt. So velocity and position are always updated as shown in line number 15 and 16 irrespective of anything. But this is not the case for non ideal candidate. Based on random number generated and SuccessProbability, non-ideal candidate is classified to facing either success or failure. Non ideal candidate will not update velocity and position and moves into halted state when he faces failure as shown in line number 25. He updates velocity and position when he faces success as shown in line number 21 and 22. Hence failure or success is not a matter for ideal candidate. But non ideal candidate will stop progress when he faces failure.

Human Poverty Particle Swarm Optimization
There are no Hybrid PSO algorithms based on Human Poverty till date. The population consists of Rich Humans and Poor Humans. Based on random number generated and RichCandidateProbability, the human is classified into either Rich or Poor. Rich Humans have enough money to move in the search space without any halt. So velocity and position are always updated as shown in line number 15 and 16 irrespective of anything. But this is not the case for poor Humans. Based on random number generated and DonationsProbability, Poor Human is classified to having enough money to move in the search space or having insufficient money. Poor Human will not update velocity and position and moves into halted state when he doesn't have enough money as shown in line number 25. He updates velocity and position when he gets donations and has enough money to travel in search space as shown in line number 21 and 22. Hence money is not a matter for Rich Human. But Poor candidate will stop progress when he did not get sufficient money to travel in search space.

Human Selection Particle Swarm Optimization
There are no Hybrid PSO algorithms based on Human Selection till date. There are 2 options to select from for Humans. Either Humans move towards local best position or they move towards global best position. Based on random number generated and HumanSelectionProbability, Humans select from 2 options available. If random number generated is less than HumanSelectionProbability then Human move towards local best as shown in line number 15. Otherwise, Human move towards global best position as shown in line number 20.

Human Safety Particle Swarm Optimization
Please see [25], to understand Human Safety Particle Swarm Optimization (HuSaPSO). The code for HuSaPSO is shown below.

Human Kindness Particle Swarm Optimization
Please see [25], to understand Human Kindness Particle Swarm Optimization (HKPSO). The code for HKPSO is shown below.

Multiple Strategy Human Particle Swarm Optimization
Please see [25], to understand Multiple Strategy Human Particle Swarm Optimization (MSHPSO). The code for MSHPSO is shown below.

Human Thinking Particle Swarm Optimization
Please see [25], to understand Human Thinking Particle Swarm Optimization (HTPSO). The code for HTPSO is shown below.

Human Disease Particle Swarm Optimization
Please see [25], to understand Human Disease Particle Swarm Optimization (HDPSO). The code for HDPSO is shown below.
Artificial Human Optimization Algorithms (AHO Algorithms) inspired by Bhagavad Gita (HBGPSO), Human Poverty (HPPSO), Human Dedication (HuDePSO) and Human Selection (HuSePSO) are proposed in this work. Ten AHO algorithms are applied on 5 benchmark functions and results obtained are shown in this work. Six AHO algorithms performed as good as PSO algorithm where as remaining four AHO algorithms didn't performed as good as PSO. HuSaPSO performed worst among all algorithms used in this work. All algorithms designed in this work performed as good as PSO. A general misunderstanding among people is that algorithms inspired by Humans will perform better than other algorithms inspired by other beings. For example, let algorithm A is inspired by Birds and Algorithm B is inspired by Humans. Then because of misunderstanding, it will lead to conclusion that Algorithm B performs better than Algorithm A because Humans are best beings and most intelligent beings on this planet. In this work, we have found that HuSaPSO inspired by Humans did not performed well even on single benchmark function where as PSO inspired by birds performed well on all benchmark functions. Our future work is to design "Human Cricket Particle Swarm Optimization (HCPSO)", "Human Farming Particle Swarm Optimization (HFPSO)" inspired by Human Cricket game and Human Farming respectively. Artificial Human Optimization Algorithms designed from scratch will also be part of our future work.