Data science is concerned with the processing and analysis of large quantities of data to create models that can be used to make predictions or otherwise support a specific goal. This process often involves the building and training of models. The specific approach to solve a problem is dependent on the nature of the problem. However, in general, the following are the high-level tasks that are used in the analysis process:
Acquiring the data: Before we can process the data, it must be acquired. The data is frequently stored in a variety of formats and will come from a wide range of data sources.
Cleaning the data: Once the data has been acquired, it often needs to be converted to a different format before it can be used. In addition, the data needs to be processed, or cleaned, so as to remove errors, resolve inconsistencies, and otherwise put it in a form ready for analysis.
Analyzing the data: This can be performed using a number of techniques including:
Statistical analysis: This uses a multitude of statistical approaches to provide insight into data. It includes simple techniques and more advanced techniques such as regression analysis.
AI analysis: These can be grouped as machine learning, neural networks, and deep learning techniques:
Machine learning approaches are characterized by programs that can learn without being specifically programmed to complete a specific task
Neural networks are built around models patterned after the neural connection of the brain
Deep learning attempts to identify higher levels of abstraction within a set of data
Text analysis: This is a common form of analysis, which works with natural languages to identify features such as the names of people and places, the relationship between parts of text, and the implied meaning of text.
Data visualization: This is an important analysis tool. By displaying the data in a visual form, a hard-to-understand set of numbers can be more readily understood.
Video, image, and audio processing and analysis: This is a more specialized form of analysis, which is becoming more common as better analysis techniques are discovered and faster processors become available. This is in contrast to the more common text processing and analysis tasks.
Complementing this set of tasks is the need to develop applications that are efficient. The introduction of machines with multiple processors and GPUs contributes significantly to the end result.
While the exact steps used will vary by application, understanding these basic steps provides the basis for constructing solutions to many data science problems.
Java and its associated third-party libraries provide a range of support for the development of data science applications. There are numerous core Java capabilities that can be used, such as the basic string processing methods. The introduction of lambda expressions in Java 8 helps enable more powerful and expressive means of building applications. In many of the examples that follow in subsequent chapters, we will show alternative techniques using lambda expressions.
There is ample support provided for the basic data science tasks. These include multiple ways of acquiring data, libraries for cleaning data, and a wide variety of analysis approaches for tasks such as natural language processing and statistical analysis. There are also myriad of libraries supporting neural network types of analysis.
Java can be a very good choice for data science problems. The language provides both object-oriented and functional support for solving problems. There is a large developer community to draw upon and there exist multiple APIs that support data science tasks. These are but a few reasons as to why Java should be used.
The remainder of this chapter will provide an overview of the data science tasks and Java support demonstrated in the book. Each section is only able to present a brief introduction to the topics and the available support. The subsequent chapter will go into considerably more depth regarding these topics.