Results for tag: classification

Linear Classifier with Softmax

The idea of this notebook is to explore a step-by-step approach to create a Linear Classifier with Softmax without the help of any third party library. We will later compare its results with two other supervised learning algorithms such as Neural Network and K-Nearest Neighbors in order to see if there’s any difference in performance and accuracy.

In practice, these Algorithms should be useful enough for us to classify our data whenever we have already made clusters (in this case color) which will serve as a starting point to train our models.

Read More

K-Nearest Neighbors Classification

The idea of this notebook is to explore a step-by-step approach to create a K-Nearest Neighbors Algorithm without the help of any third party library. In practice, this Algorithm should be useful enough for us to classify our data whenever we have already made clusters (in this case color) which will serve as a starting point to find neighbors.

Read More