Download e-book Swarm Intelligence for Multi-objective Problems in Data Mining

Free download. Book file PDF easily for everyone and every device. You can download and read online Swarm Intelligence for Multi-objective Problems in Data Mining file PDF Book only if you are registered here. And also you can download or read online all Book PDF file that related with Swarm Intelligence for Multi-objective Problems in Data Mining book. Happy reading Swarm Intelligence for Multi-objective Problems in Data Mining Bookeveryone. Download file Free Book PDF Swarm Intelligence for Multi-objective Problems in Data Mining at Complete PDF Library. This Book have some digital formats such us :paperbook, ebook, kindle, epub, fb2 and another formats. Here is The CompletePDF Book Library. It's free to register here to get Book file PDF Swarm Intelligence for Multi-objective Problems in Data Mining Pocket Guide.

IEEE CEC 12222 Special Sessions

In the second part of the chapter, an application of the ABC algorithm to colormap quantization is presented. Results of the ABC algorithm was compared to those of k-means, fuzzy-c-means and particle swarm optimization algorithms. It can be reported that compared to the k-means and fuzzy-c-means algorithms, the ABC algorithm has the advantage of working with multi-criterion cost functions and being more efficient compared to particle swarm optimization algorithm.

Foraging behavior of social creatures has always been a matter of study for the development of optimization algorithms. SMO exquisitely depicts two fundamental concepts of swarm intelligence: self-organization and division of labor. SMO has gained popularity in recent years as a swarm intelligence based algorithm and is being applied to many engineering optimization problems. This chapter presents the Spider Monkey Optimization algorithm in detail. A numerical example of SMO procedure has also been given for a better understanding of its working.

A Genetic Algorithm GA is a stochastic search method that has been applied successfully for solving a variety of engineering optimization problems which are otherwise difficult to solve using classical, deterministic techniques. GAs are easier to implement as compared to many classical methods, and have thus attracted extensive attention over the last few decades.

Evolutionary and Swarm Intelligence Algorithms

However, the inherent randomness of these algorithms often hinders convergence to the exact global optimum. In order to enhance their search capability, learning via memetics can be incorporated as an extra step in the genetic search procedure. This idea has been investigated in the literature, showing significant performance improvement. In this chapter, two research works that incorporate memes in distinctly different representations, are presented. In particular, the first work considers meme as a local search process, or an individual learning procedure, the intensity of which is governed by a theoretically derived upper bound.


  • chapter and author info;
  • ICSI Call for Special Session Proposals!
  • Summer Desserts!
  • Kundrecensioner.
  • An Analysis of Economic Size Distributions.

The second work treats meme as a building-block of structured knowledge, one that can be learned and transferred across problem instances for efficient and effective search. In order to showcase the enhancements achieved by incorporating learning via memetics into genetic search, case studies on solving the NP-hard capacitated arc routing problem are presented. Moreover, the application of the second meme representation concept to the emerging field of evolutionary bilevel optimization is briefly discussed.

Multi-objective optimization problems are common in practice. In practical problems, constraints are also inevitable. The population approach and implicit parallel search ability of evolutionary algorithms have made them popular and useful in finding multiple trade-off Pareto-optimal solutions in multi-objective optimization problems since the past two decades.

In this chapter, we discuss evolutionary multi-objective optimization EMO algorithms that are specifically designed for handling constraints. Numerical test problems involving constraints and some constrained engineering design problems which are often used in the EMO literature are discussed next. The chapter is concluded with a number of future directions in constrained multi-objective optimization area.

Our objective is to provide a comprehensive introduction to Genetic Programming GP primarily keeping in view the problem of classifier design along with feature selection. We begin with a brief account of how genetic programming has emerged as a major computational intelligence technique. Then, we analyse classification and feature selection problems in brief. We provide a naive model of GP-based binary classification strategy with illustrative examples. We then discuss a few existing methodologies in brief and three somewhat related but different strategies with reasonable details. Before concluding, we make a few important remarks related to GP when it is used for classification and feature selection.

Top Authors

In this context, we show some experimental results with a recent GP-based approach. Designing effective scheduling rules or heuristics for a manufacturing system such as job shops is not a trivial task. In the early stage, scheduling experts rely on their experiences to develop dispatching rules and further improve them through trials-and-errors, sometimes with the help of computer simulations.

In recent years, automated design approaches have been applied to develop effective dispatching rules for job shop scheduling JSS. Genetic programming GP is currently the most popular approach for this task. The goal of this chapter is to summarise existing studies in this field to provide an overall picture to interested researchers. Then, we demonstrate some recent ideas to enhance the effectiveness of GP for JSS and discuss interesting research topics for future studies.

The so-called Evolutionary Fuzzy Systems consists of the application of evolutionary algorithms in the design process of fuzzy systems. A new technique has been developed to find the solution of quadratic programming problem QPP by modelling of modified Fourier elimination technique of inequalities and concept of bounds. The technique is quite useful because the calculations are simple and takes least time then earlier existing methods.

IEEE Congress on Evolutionary Computation

The technique has been illustrated by a numerical example also. Keywords : quadratic programming problem; nature inspired elimination technique; inequalities; nonlinear objective function; constraints. To sell the produced energy,many generating companies are now forced to prepare and submit daily offers to an electricity market under uncertainty in bid prices submitted by their competitors.

To calculate the bid prices optimally and for maximizing the profit of generating company, this paper presents an optimal bidding strategy for a generating company. The company is operating in a day ahead market for single and multi-hourly uniform price multi-unit auctions. By using position update equations of SCA, the feeding characteristics of the whales are improved.

By suggested approach, the optimal solution for Market Clearing Price MCP , load dispatch and bid cost under five different capacity and price blocks are calculated. After meaningful comparison from other meta-heuristic techniques, it is observed that the profit obtained by the proposed approach is significantly higher for single-hourly and multi-hourly trading trends. In SI based algorithms agents actrnin a group and learn from each other for food foraging survivingrnetc. Teaching-learning based optimization algorithm TLBOA isrnan efficient approach of dealing with linear, nonlinear and multidimensionalrnoptimization problems established by Dr.

VenkatarnRao in Since its inception, a lot of research has been carriedrnout to make TLBOA more proficient and to apply it to differentrntypes of optimization problems. This paper presents a review ofrnTLBOA developments, applications, comparative-performance,rnand future research perspectives. Sometimes, industries face the problem due to speed, position, flexibility, reliability and higher cost. This paper presents PID controller which is simply tuned with the nature-inspired algorithm that gives better dynamic and static performance with high accuracy.

The vector control of IM includes control of magnitude and phase of each phase current and voltage. In this research paper, the field orientated control, a digital implementation which demonstrates the capability of performing direct torque control, of handling system limitations and of achieving higher power conversion efficiency is considered. The obtained outcomes are significantly better than other state-of-art algorithms available in the literature.

So far, initial ACO parameters have been set using a trial and error procedure and sometimes with some heuristics. However, these approaches lead to slow convergence and sometimes even divergence of the designed fractional order controller. Our proposed approach advocates a genetic algorithm GA based initialization of the ACO parameters for efficient and effective design of the ACO based fractional order controllers.

This brings in an element of certainty in the performance of the designed ACO controller. The controller has been validated on seven fractional order systems with comparative evaluation against: a ACO control and, b GA based fractional order control. Our GA based ACO controller is able to garner better transient response parameters rise time, peak overshoot along with an excellent steady state performance as signified by settling time, ITAE.

A flip side though, is higher computational complexity easily tackled by high performance machines of our approach which is mitigated to a certain extent by its superior performance. The deep drawing process contains many components and steps. Pots, pans for cooking, containers, sinks, automobile parts such as panels and gas tanks are among a few of the items manufactured by sheet metal deep drawing.

The Predominant failure modes in sheet metal parts deep drawing process are fracture. The prediction and prevention of fracture are extremely important in the design of tooling and process parameters in deep drawing process. Fracture or necking occurs in a drawn part, which is under excessive tensile stresses. Fractures are the important defects in deep drawing operation, which can be prevented using blank holding force. Fracture limit depends on various tooling, process and material parameters.

Firefly algorithm is one of the evolutionary optimization algorithms, and is inspired by fireflies behaviour in nature. Each firefly movement is based on absorption of the other one. This paper introduces modified particle swarm optimization technique called as APSO Adaptive particle swarm optimization and PSO for trajectory length optimization. For estimating the trajectory length of the robot, nine numbers of obstacles is selected between start and goal point in a static environment.

Lastly a comparison is established between these two approaches, to identify the approach that affords shortest trajectory length in a least computation time and shortest possible travel time.

4 editions of this work

Simulation result shows that APSO contributes towards curtail trajectory length at a lesser computational and travel time as compared to PSO. As the model is nonlinear in nature, to accomplish the desired power production level in the constant power region, an adaptive controller is implemented. It takes care of the pitch control with online estimates of the plant parameters that are susceptible to change due to disturbances.

Here, the controller design is based on the pole-placement methodology for a self-tuning controller STC. Location of the desired pair of poles is defined by the damping factor and natural frequency.

A Monte Carlo Simulation has been carried out for comparison of the algorithms. In this paper, the fine tuning of the update equations for the swarm are done based on linkage of particle motion with a electromagnetic field and also under the influence of strategic delays. The motion of a particle in a search space is confined to free space in general, however if restricted the solution under the envelope of a magnetic field, the algorithm better converges within a electromagnetic field.

The studies are then applied to the triple inverted pendulum case study which showed that stability was achieved with ease when compared to classical methods of control. The main objective of this paper is to obtain optimal PID gains of speed controller at different operating conditions. The efficient PID tuning is based on minimization of integral square error which is the objective function of this optimization problem.

At first, a two area power system is considered in which area-1 having thermal, distributed units and in area-2 includes thermal, hydel and nuclear units. Appropriate nonlinearities such boiler dynamics, governor dead band GDB and generation rate constraint GRC are considered. Finally, sensitivity of the proposed controller is investigated over a wide variation of system parameters and loading condition. For more examination of the proposed controller is also analyzed under random step load and sinusoidal disturbances. Researchers have advanced profuse algorithms by replicating the swarming behavior of different creatures.