pl-image_classification

Abstract

This is a demo ChRIS plugin app.

The image plugin runs a neural network to perform a supervised machine learning task (Classification) on the MNSIT data set.

The MNIST data set is a database of images of handwriten digits, with 60,000 training data points and 10,000 test data points.

You can learn more about the MNIST data set here: http://yann.lecun.com/exdb/mnist/

Run

• In your pwd, make two directories and name them 'in' and 'out'.
• Download the dataset from http://yann.lecun.com/exdb/mnist/ and move the

downloaded files (.gz) files to the 'in' folder.

• Using docker run

---

Assign an "input" directory to /incoming and an output directory to /outgoing

docker run --rm -v $(pwd)/in:/incoming -v $(pwd)/out:/outgoing   \
        fnndsc/pl-image_classification image_classification.py            \
        /incoming /outgoing

This will ...

Make sure that the host $(pwd)/out directory is world writable!