Machine Learning
Introduction
Machine learning algorithms is an umbrella term that covers a range of generic algorithms which falls into two main categories: supervised learning and unsupervised learning.
Supervised learning means you know the solution to a problem and you make the computer determine an acceptable solution which can then be applied to any other problem of the same type.
Unsupervised learning is basically what data mining is: it attempts to look for hidden patterns in the data.
Multivariate Linear Regression
Multivariate linear regression is a 3 step algorithm that takes in a set of data (ie. the training data) with a solution and applies weights to a number of properties in order to get as close to the answer as possible.
The 3 steps are:
- For each parameter, set its weight to 1.0
- For every piece of data, see how far off the function using the given weights are off by. The average error amount is the cost of the function.
- Repeat step 2 with different weights to minimize the cost (ie. the error)
Step 3 requires picking new weights that reduces the cost. This can be done by something called a batch gradient descent which picks values that 'walks' down the gradient. Because of this, the data that can be handled with this approach has to be linear.
Other Topics