Monday, September 26, 2022
HomeUncategorizedArtificial Intelligent Advanced Algorithms Based-IT Support Services

# Artificial Intelligent Advanced Algorithms Based-IT Support Services

There are three types of algorithms to solve the the problem of optimal bandwidth of GWR. These algorithms are the the Golden Section Search algorithm, Lightning algorithm and Heuristic Optimization Algorithm. In this survey paper, we explain these three algorithms and illuminate the working of these algorithms.

GWR is an analytical technique that considers non-stationary objects like climate, environmental conditions. GWR is a type of regression.  It is for when there are different outcomes in different places.

GWR predicts the relationship between variables, like location and the outcome variable (e.g., salaries). In this research paper progress report, we have discussed cross-validation and in the next report we will show you what it is. In this one, we will show you how to use it in a few simple steps. Get the best IT Services and Support by Stampa Solutions

## Cross-Validation:-

If we take data for training, it will never train on the validation data. But if we take the same data for both training and validation, the model may not be able to work with that data. You can use cross-validation to see how much work you need in order to train your model with this data set.

## Equal Interval Search Method:-

One of the best ways to find local maximum and local minimum is by using the Equal Interval Search Method. For the sake of simplicity, let’s just find the maximum value of function f(x), where we will know that it is a local maximum when it is in [a b].

The value of the function is increasing from point A to point M-, and the value of the function is decreasing from M+ to B. So, if there is a sudden change in the pattern of increasing or decreasing, then we can find new lower and upper bounds.

### Where GSS method is use for and how does it work:-

The GSS method is used to find the minimum or maximum from a unimodal function. Unimodal just means one of the values in a certain interval, like [0, 10]. For example, let’s say you have a function that is graph on an interval [0,10]. The GSS method would be use to find the maximum value of this function. The Equal Interval Search Method is sometimes not good enough because when it’s very small in size, it takes forever to find the maximum value of this kind of function.

## Lightning Search Algorithm:-

Optimization is the process of finding the best solution. Optimization can be good or bad, depending on what you want to do. LSA is a new and effective optimization method designed for solving real-valued optimization problems. It works just like lightning and moves along in a tree-root type way. LSA has three different search methods that it uses to find the best answer for your problem: Standard, Dynamic, and Nested Search Methods.

There are many different ways to solve this problem. Some methods come before others. They use a predefined way of solving the problem. One is the direct method and another is the gradient method. But if you have a big problem, these methods might not work because they take a lot of time and energy.

Computational intelligent algorithms like LSA are nature-inspired computational models that can address real-world problems and also solve them in an optimized way. Computational intelligent algorithms can be divided into swarm intelligence methods and evolutionary algorithms (EAs). As the name suggests that Swarm intelligent algorithms are used to reduce the mathematical complexity of complex insect and animals groups. The most popular swarm intelligent algorithms are PSO, ABC and ACO. Particle Swarm Optimization Algorithm mimics the movement of birds flocking and fish schooling.

## Artificial Bee Colony

Artificial Bee Colony is inspired by bees finding honey. Ant Colony Optimization was developed after ants found the best path from their colony to food. These algorithms are smart, but they can get trapped in local minima (find the smallest number). They also can find a suboptimal solution before it’s completed. These are common problems with genetic algorithms.

EA developed their method from natural genetic evolution. The idea is simple: the best individuals should survive and produce offspring that look like them. The goal of studying all these algorithms is to make novel LSA solve the optimization problems.

## Nature Inspired Optimization Algorithms:-

### 1) Bat Algorithm:-

This is a high-level algorithm from Xin-She-Yang. It uses the way bats use echolocation with different rates and loudness. Bats fly around randomly and they go to places with a different sound, too. Bats change how they use sound when they find food. They can change the frequency and loudness of sounds, and also the rate at which they send out pulses (sound waves). Many researchers have found these things about bats and made strategies like fuzzy logic, chaotic sequence and levy flight concept etc., to help them.

### 2) Firefly Algorithm:-

The firefly algorithm is an optimization method which calculates the best solution. It involves two major issues, the variation of light intensity and finding a way to make it attractive for other fireflies. The attractiveness between the fireflies is based on how close they are to each other and how much light they absorb. As more fireflies find a solution, their movements will be calculated by the the center point of where they started from.

### 3) Teacher Learning Optimization Algorithm:-

This is an algorithm for teachers and students. Teachers tell students what they know, and the students learn it. The first phase is the teacher phase, and the other is the student or learner phase. During this process, teachers teach their lessons to students by telling them what they know about that topic. The student phase is the part of the algorithm that helps learn. It can be hard if there is only one teacher, since they might teach too much. This could cause the algorithm to not be very well optimized.

### 4) Social Spider Algorithm:-

SSO is an algorithm that considers two search spiders and their behavior. One spider is male and the other spider is female. The male spider goes to colonies that are close, while the female spider searches for colonies farther away. The SSO algorithm has a problem with premature convergence and incorrect exploration.

### Lightning Search Algorithm:-

This solution is take from an algorithm make by A. Breakopen named “Fast particles as initiators of step leaders in CG and IC lightning” in 2012. The algorithm involves fast projectiles, which are like binary tree structures. They make two leader tips instead of the traditional step leader mechanism as described above.

### Projectile and Step Leader Propagation:-

Different molecules of oxygen, hydrogen. And nitrogen can be found near thunderclouds. If the thundercloud is cold, then water molecules will freeze and some of them won’t fit in the form. When they break out, atoms of hydrogen and oxygen separate with high speed and go in random directions as projectiles. The projectiles show how to solve a problem where there might not be an answer.

## Heuristic Optimization Algorithm:-

### Introduction:-

These days, there is a trend to get maximum output with minimal effort. There are some traditional techniques use to solve optimization problems–linear programming, nonlinear programming, and dynamic programming. But now we use heuristic optimization algorithms for the optimization of the use problems.

Linear Programming, Nonlinear Programming, and Dynamic Programming are ways of finding the best solutions to problems. But they can’t be use when there are real-world problems. In DP, when you add more variables in a recursive function in a computer program, the number of times the program will run increases.

This can make it hard for the computer to find the answer because its memory is limited. The algorithm may not found. What it needs and just get stuck on something that isn’t very good. You need to be careful about setting your initial values so that you don’t get stuck on “local optima”.

## Conclusion:-

In this article, we have talked about three ways to find the best bandwidth for a geographically weighted regression. We have talked about Golden Section Search, Lightning Search and Harmony Algorithm. In the end, we show how to make an Adaptation of the GSS Algorithm.

We found that the form three algorithms, major algorithm is the best. But we found out that GSS + Division Algorithm works better than other algorithms, and you will find out more about it in the next paper. After we compare all of these algorithms, we will show how each one works with data set.

RELATED ARTICLES