Classification

The problem of classification is an interesting one.

Some concepts you must understand

• Confusion Matrix
• Precision-Recall

Classification problems usually depend on a threshold to classify something as true or false.

In Stanford CS231n, it was a Classification problem, where you used the SVM loss to predict the class.

For example, for Object Detection, there is the IoU metric.

When you do classification, you don’t use Mean Squared Error. You can use for example a Mean Squared Error. You can use for example a Support Vector Machine

Seems like they use Softmax Function oftentimes, with a Softmax Function oftentimes, with a Softmax Classifier

More specific Problems:

• Image Classification