/machine-learning

Repo containing deep learning models and algo

Primary LanguageJupyter Notebook

Machine Learning Algorithms implemented

  1. mnist_lr -This notebook containts the implementation of linear regression algorithm to classify MNIST coded from scratch in numpy.
  2. basic/mnist_numpy.ipynb - This notebook containts a simple 2 layer neural network to classify MNIST coded from scratch in numpy.
  3. pca_mnist.ipynb - Used tensorflow keras to implement a classification algorithm for the mnist dataset which further involved dimensionality reduction using PCA
  4. iiitpets.ipynb - Used tf estimators to create a simple classifier to classify various categories in Oxford-IIIT Pets Dataset
  5. omniglot_oneshot.ipynb - Used Keras to implement a Siamese Network along with triplet loss to implement oneshot learning on omniglot dataset
  6. gan.ipynb - Used tf1 to implement a Generative Adversarial Network (GAN) to generate MNIST from noise
  7. denoising_vae.ipynb - Usef tensorflow 2, custom model to implement a denoising variational auto-encoder to denoise cifar-10 dataset from shot noise