MNIST - Handwritten Digit Recognition

By for May 4, 2018

Report Abuse
This model predicts handwritten digits using a convolutional neural network (CNN) using MNIST (Modified National Institute of Standards and Technology) dataset.
This model predicts handwritten digits using a convolutional neural network (CNN). **Model size: 26 kB** **Dataset** The model has been trained on the popular MNIST [dataset.][1] **Source** The model is trained in CNTK following the tutorial [CNTK 103D: Convolutional Neural Network with MNIST][2]. Note that the specific architecture used is the model with alternating convolution and max pooling layers (found under the "Solution" section at the end of the tutorial). **Model input and output** **Input** * Input image of the shape (1x28x28) **Output** * Output is a (1x10) array **Pre-processing steps** Resize the input image to a (1x28X28) array of type float32. **Post-processing steps** Route the model output through a softmax function to map the aggregated activations across the network to probabilities across the 10 classes. **Sample test data** Sets of sample input and output files are provided in .npz format (test_data_*.npz). The input is a (1x28x28) numpy array of an MNIST test image, while the output is an array of length 10 corresponding to the output of evaluating the model on the sample input. **License** [MIT][3] [1]: http://yann.lecun.com/exdb/mnist/ [2]: https://github.com/Microsoft/CNTK/blob/master/Tutorials/CNTK_103D_MNIST_ConvolutionalNeuralNetwork.ipynb [3]:https://github.com/onnx/models/blob/master/LICENSE