Practical Machine Learning: A New Look at Anomaly Detection

Chapter 7. Anomaly Detection for the Future

The present is an exciting time for those who are interested in machine learning. The surge of interest in extracting value from growing data sets at large scale has opened up the field for new applications of basic and novel techniques, as well as opened the job market in order to fill the sudden new demand for data scientists and developers experienced with machine learning. The rapid expansion of the use of machine learning in mainstream business operations also means there is increasing importance in designing new, practical approaches that are both approachable and very effective.

These changes also raise the stakes for being able to effectively communicate about these highly technical topics between teams with very different areas of knowledge. This need underlines the usefulness of learning the fundamental concepts and basic approaches to machine learning in order to discuss them in a comprehensible way with decision makers for business solutions, individuals possessing domain knowledge relevant to your project, technical practitioners who more often think in terms of math and code, and newcomers to data science. You must develop the habit of being able to speak of fundamental concepts and methods, using clear and widely understood terms, in order to foster excellent exchanges between these different groups.

With these goals in mind, we also want to look toward the future. Our prediction is that anomaly detection is certainly going ...