Ever wondered what classification models do?
In machine learning, there are many different types of models, with various outcomes. When it comes to classification or statistical classification, the model is trying to identify two or more determined classes, i.e. Apples and Bananas, and classify them accordingly. Usually, these models have been trained using a training set.
What You'll Learn
- Introduction to classification models
- The use of classification models in machine learning
Welcome to this short introduction to classification models.
What are classification models? In machine learning there are many different models all with different types of outcomes.
Classification models are machine learning models that predict a class type outcome. In other words a classification model predicts any kind of category or class such as apples and bananas.
A classification model uses attributes of a person or any kind of entity to predict the entities class, for example Class A might be apples and Class B might be bananas. The attributes of apples and bananas could be their shape their dimensions and their color. These data points could be used to predict the class outcome of it likely being an apple or a banana differentiating apples from bananas based on their own unique attributes this means the model learns that certain attributes belong to a certain categories or classes for example if it’s colored yellow is six to eight inches long one to two inches wide and it’s crescent-shaped then these attributes are more likely to belong to a banana than an apple. The model makes a prediction that given these attributes the fruit is likely to be a banana similarly if it purrs has fur and whiskers and is found in every corner of the internet then it’s likely a cat; if a croaks, has feathers and wings, and is found on farms, it is likely a rooster.
A classification model learns that these attributes belong to a certain categorical outcome in a supervised way where it directly maps the data points to a class label. The class label can be binary such as positive or negative, whether a disease is present or not, whether the customer is a returning customer or not, or whether the job applicant was a success or fail or the class label could be multiple classes such as easy, intermediate, and advanced level in a game, or all types of fruits from peaches, oranges, and kiwi, not only apples and bananas some key algorithms used in classification models include decision trees, naïve bayes, support vector machines, and neural networks, which you can learn about these in future videos. They all take different approaches to predicting a class outcome. And that quickly sums up classification models for you! Thanks for watching, give us a like if you found this useful, or you can check out our other videos at online.datasciencedojo
Raja Iqbal - Raja Iqbal is a data scientist, a passionate educator, and an internationally recognized speaker on all things data science. He is the Founder and Chief Data Scientist at Data Science Dojo. Prior to Data Science Dojo, Raja worked at Microsoft in a variety of research and development roles involving machine learning and data mining at very large scale. Raja has a Ph.D in Computer Science from Tulane University with a focus on machine learning and data mining.