Handbook of Machine Learning

The following sections are included:

• Preface
• About the Authors
• Acknowledgments
• Contents

This chapter introduces this book, which is a continuation of Volume 1. This book addresses three key issues and these are optimization, rational decision-making and applications within the context of artificial intelligence (AI). This chapter further describes concepts in AI, optimization and decision-making. It concludes by outlining the summary of the book.

This chapter describes the Nelder–Mead (NM) simplex method, the Broyden–Fletcher–Goldfarb–Shanno (BFGS) and conjugate gradient (CG) methods. They are then applied to estimate parameters in a finite element model given the experimental data. The NM method performed better than the other methods.

This chapter introduces the basic working principles of genetic algorithm (GA). This method was tested on typical used products collection problem in the context of remanufacturing. It was observed that genetic algorithm was able to find the shortest travel plan, while keeping the fuel consumption rate at the lowest possible level. Furthermore, GA was used to update the finite element (FE) model to better reflect the measured data. The results obtained indicated that GA was able to successfully update the FE model resulting in better reflection of the measured data.

This chapter describes the particle swarm optimization (PSO) method. Its evolution, theory and applications are explored. Examples on how it has been applied for missing data estimation, operating system scheduling as well as finite element updating are explored. The results obtained showed that the particle swarm optimization gave good results.

This chapter describes simulated annealing (SA) and its theory. Simulated annealing is applied for solving Sudoku as well as to update the finite element model (FEM). The variant of simulated annealing implemented is the quantum simulated annealing. The results indicated that simulated annealing was able to solve the Sudoku puzzle and also to update an FEM.

We describe the theory and applications of the response surface method (RSM). We constructed the RSM by estimating the finite element (FE) surface response equation using the multi-layer perceptron (MLP). We used genetic algorithm (GA), particle swarm optimization and simulated annealing (SA) to optimize the surface response equation. We applied the RSM to estimate the parameters of FE models of a simple and an unsymmetrical H-shaped structure given the measured data.

This chapter presents the ant colony optimization (ACO) technique. It is based on the behavior of ants and how they are able to identify an optimal path between the food source and the nest. The ants operate by depositing pheromones on their paths. The trail with the strongest pheromones is the optimal path that is followed by most of these ants at an equilibrium point. We apply the ACO technique to the condition monitoring and missing data estimation problems.

This chapter is about the bat and firefly algorithms (FA). The bat algorithm (BA) is a metaheuristic swarm intelligence technique based on the echolocation trait of bats. Bats possess incredible capabilities with their ability to hunt for prey in the dark using sonar and the Doppler effect. The firefly algorithm is a nature-inspired metaheuristic algorithm based on the flashing patterns and behavior of fireflies. We apply the algorithms to the missing data estimation problem.

This chapter is on artificial immune system (AIS) which is a biological metaphor for the human immune system, just like a neural network is based on the way the human brain functions. We describe some basic concepts of the AIS algorithm and the negative selection algorithm which is one of the approaches used to implement the AIS. We apply the algorithm to the missing data problem.

This chapter is on invasive weed optimization (IWO) and Cuckoo search (CS). IWO algorithm is a method which mimics the invasion trait of weeds. In general, weeds are defined as plants that grow in an area where they are not wanted. Specific to horticulture, the term weed refers to a plant whose growth properties poses a menace to plants that are cultivated. Weeds portray fascinating traits, such as adaptivity and robustness. CS algorithm is a population-based metaheuristic technique based on the brood parasitism trait of certain species of cuckoo birds. We apply the algorithms to the missing data estimation problem.

This chapter is on the Decision Tree (DT) algorithm. Decision trees are predictive models that make use of a set of bipartite rules that lead to a desired output being obtained. It can be used for categorical or continuous output values. Random forests are an ensemble learning method for classification and regression, among others, that operates by constructing multiple decision trees at training time and outputting the class that is the mode of the classes or mean prediction. We apply the algorithms to the missing date estimation problem.

A hybrid has two or more components that produce the same or better results, for example, a vehicle powered by both an electric motor and an internal combustion engine as sources of power for the drive train. We apply the hybrid algorithm of the particle swarm optimization (PSO) and Nelder–Mead (NM) simplex method for finite element model (FEM) updating. The results observed showed that on average the hybrid gave results that were more accurate, followed by the PSO and then the NM simplex method.

This chapter introduces economic and financial modeling based on artificial intelligence (AI). It explores subjects such as data representation and modeling. It studies several types of AI-based modeling tools such as neural networks, support vector machines (SVMs), rough sets (RSs) and incremental learning within the context of economic and financial modeling.

This chapter is on condition monitoring. The condition monitoring domain in which the data is visualized is discussed, and in particular the time, modal, frequency, and time–frequency domains. The generalized condition monitoring framework which includes the data acquisition device, data analysis device, feature selection Device and decision-making device is also presented. Techniques for using these decision-making devices are introduced. These are the finite element models, correlation based methods and computational intelligence methods.

This chapter studied unbounded, bounded and flexibly bounded rational decision-making. It explored whether rationality can be measured. It also explored whether machines are more rational than human beings and whether markets are rational. Furthermore, this chapter studies the theories of bounded and flexibly bounded rationality. It uses two concepts to study these two theories and these are condition monitoring as well as HIV problems. In conclusion, flexibly bounded rationality moves the bounds of rationality for better decision-making.

This chapter gives a summary of the topics that were discussed in this book. These include classical optimization techniques such as the Nelder–Mead (NM) simplex method, Broyden–Fletcher–Goldfarb–Shanno (BFGS) and conjugate gradient methods. It also describes the genetic algorithm (GA), simulated annealing and particle swarm optimization (PSO) methods. Furthermore, it describes the response surface method (RSM) as well as hybrid methods. Also presented in this book are advanced topics such as economic and financial modeling based on artificial intelligence (AI), unbounded, bounded and flexibly bounded rational decision-making.

The following section is included:

• Index