Neural Network Project (PV021) -- Deep Learning from Scratch

Project implements a neural network in low-level manner (without use of advanced libraries or frameworks) using .NET 8 supporting native AOT compilation, but JIT compilation seems to be more performant out-of-the box.

Project implements MLP architecture of Neural Network in its core to solve FashionMNIST problem with accuracy of 88%.

HOW TO RUN

First make the script executable, by setting proper permission

chmod u+x run.sh

Finally, run the whole neural-network training + evaluation Beware that this will take a while, depending on your machine performance. (around 15 mins on i7-13700H CPU + 32GB RAM)

./run.sh

OUTPUT

• Your solution must output two files to the root project directory: (next to example_test_predictions.csv file):
  • train_predictions.csv - your network predictions for the train set.
  • test_predictions.csv - your network predictions for the test set.
• The format of these files has to be the same as the supplied training/testing labels:
  • One prediction per line.
  • Prediction for i-th input vector (ordered by the input .csv file) must be on i-th line in the associated output file.
  • Each prediction is a single integer 0 - 9.

DATASET

Fashion MNIST (https://arxiv.org/pdf/1708.07747.pdf) - a modern version of a well-known MNIST (http://yann.lecun.com/exdb/mnist/). It is a dataset of Zalando's article images ‒ consisting of a training set of 60,000 examples and a test set of 10,000 examples. Each example is a 28x28 grayscale image, associated with a label from 10 classes. The dataset is in CSV format. There are four data files included:

• fashion_mnist_train_vectors.csv - training input vectors
• fashion_mnist_test_vectors.csv - testing input vectors
• fashion_mnist_train_labels.csv - training labels
• fashion_mnist_test_labels.csv - testing labels