All Courses
AI Career Planner
1:1 Coaching/Mentoring
Job Board
SQL
HTML/CSS/JS
Coding
Settings
LogoutThe most straightforward of the machine learning models is supervised learning.
In the supervised model, learning requires building a function that can be trained using a training data set and then applied to anonymous data to achieve some level of predicted performance. The goal is to generalize the function to data it has never encountered before. You build and evaluate a mapping function using supervised learning (see image below ).
In the first phase, you partition a data set into two types of samples: training data and test data. A test vector (the inputs) and one or more known target output values are present in both training and test data.
You use the training data set to train the mapping function until it achieves a certain level of performance (a metric for how accurately the mapping function maps the training data to the associated desired output). This happens with each training sample in supervised learning when the error (actual vs. desired output) changes the mapping function.
The trained mapping function is then tested against the test data in the next phase. The test data hasn’t been used in training, and it’s a helpful indicator of how effectively the mapping function generalizes to new data.
Support vector machines and nave Bayes are two techniques that fall within the supervised learning umbrella.