I’ve been reading a little about deep learning, and a problem typically presented to approach using neural networks is handwritten digit recognition. The data set I’ve seen repeatedly in this context is the Modified National Institute of Standards and Technology (MNIST) database.
MNIST contains 70,000 samples of handwritten digits that are divided into training and test sets.
I tend to want to see things visually, so I created two Observable notebooks to view the digits in the MNIST database. There’s probably already great tools available for this, but I felt like writing some code to do it myself.
Since they are Observable notebooks, it all happens in your web browser. Beyond that, there are no dependencies.
- MNIST Browser – view samples grouped by digit
- MNIST Viewer – view a single sample in the database using its index
