This folder contains various examples of training models on known datasets with EdgeLearning. Based on the ENABLE_MLPACK configuration flag, the high-level api will perform operations with Mlpack framework.
In order to execute the examples you need to download the desired dataset as described in the data folder.
A brief description of the examples in this folder follows:
Create a simple model with 2 hidden layers trained on dummy data allocated with randomly values in the range [-1,+1] with an user defined classification function that construct the labels.
Create a simple model with 2 hidden layers trained on dummy data allocated with randomly values in the range [-1,+1] with a set of user defined functions that represent the result to predict.
Create a custom loss layer with a user defined cost function (e.g. mean absolute error) and try the created loss layer on a simple model with 2 hidden layers trained on dummy data allocated with randomly values and with a set of user defined functions that represent the result to predict.
Create a 5 layers model that solve the classification problem of handwritten digits images from 0 to 9 proposed by the MNIST dataset. In order to try this example you need to download the MNIST dataset and put in the data directory.