From the course: Introduction to Data Science

Unlock this course with a free trial

Join today to access over 24,700 courses taught by industry experts.

Defining prediction for data science

Defining prediction for data science

From the course: Introduction to Data Science

Start my 1-month free trial Buy for my team

Defining prediction for data science

“

An important aspect of data science is discovering what data can tell us about the future. For example, based on a person's social media profile, which upcoming movies will likely interest them? What does climate and environmental data say about temperatures a few decades from now? How can a database of past clinical trial reports be used to estimate the risks of a new clinical trial? Answering questions like these involves making predictions. In this lesson, I'll identify two major types of prediction tasks the data scientists work on. The first is classification. Classification is about predicting the value of a categorical variable. For example, predicting whether an image shows a dog or a cat. The variable you want to predict is categorical, with dog and cat being the possible categories, in other words, classes. Now, there are many techniques the data scientists have developed for approaching classification. One method is K-nearest neighbors, which uses the distance between the…

Contents