Introducing machine learning with MATLAB

So far, we have learned what machine learning algorithms do; we have also understood how to recognize the different types, how to locate the right solution for our needs, and finally how to set a proper workflow. It's time to learn how to do all this in the MATLAB environment.

Solving machine learning problems becomes extremely easy with the use of the tools available in the MATLAB environment. This is because MATLAB is a strong environment for interactive exploration. It has numerous algorithms and apps to help you get started using machine learning techniques. Some examples include:

• Clustering, classification, and regression algorithms
• Neural network app, curve fitting app, and Classification Learner app

MATLAB is a software platform optimized for solving scientific problems and design. In MATLAB, calculation, visualization, and programming are integrated in an easy-to-use environment, where problems and solutions are expressed in familiar mathematical notation.

The name MATLAB is an acronym of the term matrix laboratory. MATLAB was originally written to provide easy access to software of matrices; then it evolved in the years to come, thanks to numerous user inputs. The MATLAB programming language is based on matrices that represent the most natural way to express computational mathematics. Its desktop environment invites experimentation, exploration, and discovery. The integrated graphics are easy to view and provide an in-depth understanding of the data.

The MATLAB desktop is shown in the following screenshot:

MATLAB is also characterized by the presence of specific solutions to application problems called toolboxes. Very useful for most users, MATLAB toolboxes represent solutions for many practical problems and provide the basis for applying these instruments to the specialized technology. These toolboxes are collections of MATLAB functions (referred to as M-files) that extend the MATLAB environment in order to solve particular classes of problems.

MATLAB has two specific toolboxes for processing machine learning problems. They are the Statistics and Machine Learning Toolbox and Neural Network Toolbox. While the first solves machine learning problems through statistical techniques and algorithms most widely used in this field, the second is specific to ANNs. In the following sections, we will analyze in detail the features of these tools.