Chapter 9. Databases and Neural Networks
In this chapter, we will look at and define Artificial Neural Network (ANN) and draw data from a data developer's knowledge of data, databases, and data models to help him or she understand the purpose and use of neural networks, and why neural networks are so significant to data science and statistics.
We have organized the information in this chapter into the following key areas:
- Definition of a neural network
- Relating a neural network model to a database model
- Looking at R-based neural networks
- Use cases
Today, if you ask any data scientist about the statistical methods, (or even a few) you will most likely discover that there are two most well-known statistical methods used within the practice of data science and the statistics industry today for predictive modeling. We introduced these two methods in Chapter 6, Database Progression to Database Regression.
These two methods are as follows:
- Linear regression
- Logistic regression
The linear regression method is probably considered to be the classic or most common starting point for problems, where the goal is to predict a numerical quantity. The Linear Regression (or LR) model is based on a linear combination of input features.
The logistic regression method uses a nonlinear transformation of this linear feature combination in order to restrict the range of the output in the interval [0, 1]. In doing so, it predicts the probability that the output belongs to one of two classes. Thus, it is a very well-known technique for...
In this chapter, we defined neural networks and, from a data developer's knowledge of databases and data models, grew to understand the purpose and use of neural networks and why neural networks are so important to data science. We also looked at an R-based ANN and listed some popular use case examples.
In the next chapter, we will introduce the idea of using statistical boosting to better understand data in a database.