2. FILE: Features vs Category (csv or txt)

This is the standard format where you list your columns to contain Category/Labels (classified outputs) followed by columns of Features (inputs).

Here is a simple example that has 4 types of output, and 8 inputs

Output class 1 is labeled as 1 0 0 0

Output class 2 is labeled as 0 1 0 0

Output class 3 is labeled as 0 0 1 0

Output class 4 is labeled as 0 0 0 1

Each row defines the classified output with its corresponding inputs. There are 8 inputs that may contribute to the classification of the outputs.

This demo file only shows one training sample for each class, in the real world application you would have many more samples for each class.

Here is a screenshot of the MNIST training dataset, it contains 60,000 training samples.

When loading this type of data into DLHUB, select the text file and press "Open".

Loaded file will be reviewed as below: