Essential Machine Learning

So far, in previous chapters, we went through the various ETL processes available in GCP. In this chapter, we will start our journey of machine learning and deep learning through the following topics:

• Applications of machine learning
• Supervised and unsupervised machine learning
• Overview of major machine learning techniques
• Data splitting
• Measuring the accuracy of a model
• The difference between machine learning and deep learning
• Applications of deep learning

Applications of machine learning

Machine learning encompasses a set of techniques that learn from historical data. Based on the patterns learned from historical data, the machine learning technique predicts the probability of an event happening on a future dataset. Given the way in which machine learning works, there are multiple applications of the set of techniques. Let's explore some of them in the following sections.

Financial services

Some applications in the field of finance are as follows:

• Identifying the riskiness of a loan/credit card applicant
• Estimating the credit limit of a given customer
• Predicting whether a card transaction is a fraudulent transaction
• Identifying the customer segments that need to be targeted...

Supervised and unsupervised machine learning

Supervised machine learning constitutes the set of techniques that work towards building a model that approximate a function. The function takes a set of input variables, which are alternatively called independent variables, and tries to map the input variables to the output variable, alternatively called the dependent variable or the label.

Given that we know the label (or the value) we are trying to predict, for a set of input variables, the technique becomes a supervised learning problem.

In a similar manner, in an unsupervised learning problem, we do not have the output variable that we have to predict. However, in unsupervised learning, we try to group the data points so that they form logical groups.

A distinction between supervised and unsupervised learning at a high level can be obtained as shown in the following diagram:

In...

Overview of machine learning techniques

Before going through an overview of the major machine learning techniques, let's go through the function that we would want to optimize in a regression technique or a classification technique.

Objective function in regression

In a regression exercise, we estimate the continuous variable value. In such a scenario, our predictions can be lower than the actual value or higher; that is, the error value could be either positive or negative. In such a scenario, the objective function translates to minimizing the sum of squared values of the difference between the actual and predicted values of each of the observations in the dataset.

In mathematical terms, the preceding is written as...

Summary

In this chapter, we understood the major difference between supervised and unsupervised learning and got an overview of the major machine learning algorithms. We also understood the areas where deep learning algorithms shine over traditional machine learning algorithms, through examples of text and image analysis.