Conclusion

Introduction

With ever increasing amounts of data becoming available there is enough reason to believe that smart data analysis will become even more pervasive as a necessary ingredient for technological progress. Over the past two decades Machine Learning has become one of the main-stays of information technology rather part of our life. The purpose of this book is to provide the reader with some specific applications of machine learning over the vast range of applications which have at their heart a machine learning problem.

This book is composed of eleven chapters assessing current state of the art of machine learning, from symbolic concept learning and conceptual clustering to case-based reasoning, neural networks, deep learning, and genetic algorithms. Edited chapters introduces the reader to innovative applications of machine learning techniques in the fields of biometric system, urban sciences, heart disease prognosis, software reliability prediction, data mining, knowledge discovery, computational intelligence, human language technology, user modeling data analysis & discovery sciences.

Over the past few years machine learning has made its way into various areas of administration, commerce, and industry, in an impressive way. Data mining is the most popular widely known demonstration of this phenomenon, complemented by less publicized applications of machine learning, such as adaptive systems in various industrial settings, financial prediction, medical diagnosis, and the construction of user profiles for WWW-browsers. This transfer of machine learning approach from the research labs to the “real world” has caused increased interest in learning techniques, dictating further effort in informing people from other disciplines about the state of the art in machine learning and its uses. The objective of this book is to provide the reader with sufficient information about the research oriented capabilities of machine learning methods, as well as ideas about how the user could make use of these methods to solve real-world problems.

The research issues addressed in book chapters include the relationship of machine learning to knowledge discovery in databases, software reliability prediction, the handling of noisy data, and the modification of the learning problem through feature selection algorithms for classification and clustering.

The first chapter of the book introduces the reader to innovative applications of machine learning and explore wide range of applications, from data mining in finance, marketing, and economics to learning in human language technology and user modeling. Some basic terminology of machine learning used in this chapter described as below:

More recently, in 1997, Tom Mitchell gave a well-posed definition that has proven more useful to engineering types:

For example if you want your program to predict, for example, traffic patterns at a busy intersection (task T), you can run it through a machine learning algorithm with data about past traffic patterns (experience E) and, if it has successfully “learned”, it will then do better at predicting future traffic patterns (performance measure P).

Arthur Samuel in 1959:

The highly complex nature of many real-world problems, though, often means that inventing specialized algorithms that will solve them perfectly every time is impractical, if not impossible. Some specific examples of machine learning can be related to classification and prediction, market basket analysis, likelihood of a pattern, object recognition, sequence matching etc. All of these problems are excellent targets for a Machine Learning (ML) project, and in fact ML has been applied to each of them with great success.

The process of feature selection in selecting a best subset of features, among all the features that are useful for the learning algorithms may be summarized as:

• •

  To provide faster and more cost effective models by reducing the size of the problem and hence reducing computational time and space required to run classifiers.

• •

  To improve the performance of the classifiers, firstly by removing noisy or irrelevant features secondly by reducing the likelihood of over fitting to noisy data. So the basic objective of feature selection algorithms to improve the performance of the classifier, i.e. prediction performance in the case of classification and better cluster detection in the case of clustering.