Classification and Regression with the BigML Dashboard
6.1 Introduction
There are many Machine Learning problems that can be solved using Supervised learning Machine Learning techniques. These techniques solve problems that require the prediction of an output variable (Objective Field) given a number of input variables (input Fields). These problems can be classified into two groups: Classification problems if you need to predict a category (label or class) or Regression problems if the output is a continuous value (a real number).
Classification and regression problems can be solved using multiple Machine Learning methods in BigML, such as models, ensembles, logistic regressions, and deepnets. These methods are explained in Chapter 1 , Chapter 2 , Chapter 4 , and Chapter 5 respectively. Depending on the problem you are trying to solve and the data available, some techniques may perform significantly better than others. Fusions combine these Machine Learning models and average their predictions to balance out the individual weaknesses of the single models and yield a better performance. Fusions are based on the same “wisdom of the crowds” principle as ensembles under which the combination of multiple models is often more performant than any of its individual models. The component models have to be as accurate and diverse as possible. See section 6.2 for more details.
This chapter contains a comprehensive description of BigML’s fusions including how they can be created (section 6.3 ), all configuration options available (section 6.4 ), and the different visualizations provided by BigML (section 6.5 ). See section 6.6 for an explanation of how fusions can be used to make predictions. You can also export your fusions in different formats to make local predictions faster at no cost (subsection 6.7.1 ). The process to evaluate your fusions’ predictive performance in BigML is explained in a different chapter (Chapter 7 ).
On BigML, the third tab of the main menu of the Dashboard allows you to list all your available fusions. The fusion list view (Figure 6.1 ), details the Name, the objective field Type (classification or regression), the Objective (Objective Field name), Age (time elapsed since it was created), and number of evaluations, predictions, and batch predictions that have been created using that fusion. The search menu option in the top right corner of the fusion list view allows you to search your fusions by name or ID (using the syntax “id:” followed by the fusion ID). You can also search a fusion by the parameters used to create it by typing in the search box the syntax “config:” followed by the parameters you are looking for.
![\includegraphics[]{images/fusions/fusion-listings}](images/img-0565.png)
By default, when you first create an account at BigML, or every time that you start a new Project, your list of fusions will be empty. (See Figure 6.2 .)
![\includegraphics[]{images/fusions/empty-fusion-listings}](images/img-0566.png)
Finally, in Figure 6.3 you can see the icon used to represent a fusion.
![\includegraphics[width=2cm]{images/fusions/fusion-icon}](images/img-0567.png)