IFD:EAI SoS21/course material/Session 4: Programming the Classifier Part1: Difference between revisions
From Medien Wiki
Line 25: | Line 25: | ||
There are a few new elements that we added to our code this week. We added a new class called 'KMeans'. Inside this class will be the guts of our classifier. To be able to successfully classify a new point (a sound), the classifier actually only needs to know the <code>centroid of each class</code> and a way to <code>find the closest centroid</code> to our new point. | There are a few new elements that we added to our code this week. We added a new class called 'KMeans'. Inside this class will be the guts of our classifier. To be able to successfully classify a new point (a sound), the classifier actually only needs to know the <code>centroid of each class</code> and a way to <code>find the closest centroid</code> to our new point. | ||
As a first step the classifier needs to calculate the centroids. To do that we feed it with the following data: | |||
* | * all our given points (=measurements of sounds) | ||
* for every point | * for every point we tell the classifier to which class it belongs (=classLabels) | ||
=Homework= | =Homework= |