Icon

Image_​Classification_​MNIST

Simple CNN for Image Classification

This workflow trains a simple convolutional neural network (CNN) on the MNIST dataset using Keras.

The enclosed pictures are from the MNIST dataset (http://yann.lecun.com/exdb/mnist/).




Define the network structure Read and preprocess thetraining and test data Train and apply model Extract prediction and evaluate trained network Please execute these two nodes first to unzip the training and test images arg maxMNIST imagesextract predicted digitInput: 28, 28, 1Filters: 32Kernel: 3, 3Pool size: 2, 2 Units: 100 Activation: ReLUUnits: 10Activation: SoftmaxMNIST ImagesUnzip training imagesUnzip test images Many to One Prepare test data Column Expressions Keras Input Layer Keras Convolution2D Layer Keras Max Pooling2D Layer Keras Flatten Layer Keras Dense Layer Keras Dense Layer Keras NetworkLearner Keras NetworkExecutor Scorer String To Number Preparetraining data Unzip Files Unzip Files Define the network structure Read and preprocess thetraining and test data Train and apply model Extract prediction and evaluate trained network Please execute these two nodes first to unzip the training and test images arg maxMNIST imagesextract predicted digitInput: 28, 28, 1Filters: 32Kernel: 3, 3Pool size: 2, 2 Units: 100 Activation: ReLUUnits: 10Activation: SoftmaxMNIST ImagesUnzip training imagesUnzip test images Many to One Prepare test data Column Expressions Keras Input Layer Keras Convolution2D Layer Keras Max Pooling2D Layer Keras Flatten Layer Keras Dense Layer Keras Dense Layer Keras NetworkLearner Keras NetworkExecutor Scorer String To Number Preparetraining data Unzip Files Unzip Files

Nodes

Extensions

Links

Download

To use this workflow in KNIME, download it from the below URL and open it in KNIME:

Download Workflow
Created by: marcel
Created at: 2017-07-05
On NodePit since: 2020-12-17
Last update: 2024-11-17
Created with KNIME version: v4.2.2
Tags: deep learningKerasimage classification

Deploy, schedule, execute, and monitor your KNIME workflows locally, in the cloud or on-premises – with our brand new NodePit Runner.

Try NodePit Runner!