Hands-On ROS for Robotics Programming

Core concepts in ML

Before going into the use case of object recognition in images, let's take a much simpler example, the prediction of the price of a house as a function of several independent variables: area, number of rooms, distance to the center of the city, population density, and more.

First, to have a working ML algorithm, we need an underlying model that, when fed with input data, can produce a prediction. The data has to be supplied according to the features, that is, independent variables, that we have selected for our model. Then, establishing the correspondence with our simple example, we can explain the several concepts involved in an ML problem:

• The algorithm is the computation as a whole, specified as a sequence of instructions ...