PyTorch-Scholarship-Challenge

Got Accepeted for Phase 2

Day 1 | Course Introduction

Learned Classification Problem, Linear Boundaries, Higher Boundaries and Perceptron. Perceptron is pretty amazing it worked same as a Neuron.

Day 2 | Entropy

Learned Error Function, Log-Loss Error Function and cross entropy. Among these cross entropy is the best way to get prediction. Ipython Notebook for Soft max and cross-Entropy and Perceptron

Day 3 | Gradient Descent

Learned about gradient Descent and continuous perceptron. Ipython Notebook Gradient Descent

Day 4 | First Neural Network

Learned about the Neural Network Architure, Backpropogation, made my first neural network, Regularization and different activation function. The thing about neural network is that it is a combination of linear models and the maths behind every function is beautiful. Ipython Notebook for Student Admissions

Day 5 | Neural Network using Pytorch

Build a neural network using Pytorch. Ipython Notebook for Neural Network

Day 6 | Neural Network Architectures

Build Neural Network Architecture and trained the model. Ipython notebook for Neural Network and Training Neural Networks.

Day 7 | Classifying Fashion MNIST

Build a simple Neural Network from scratch to detect the type of clothes. I got to know that with very simple model you can also get goot accuracy. Ipython Notebook of Fashion MNIST

Day 8 | Dogs vs Cats

Build a simple Neural Network from scratch to detect that the image is of cat or dog. Learned about Inference and Validation and how to save and load a trained model. Also learned how to load image files on pytorch. Ipython Notebook of Dogs vs Cats, Inference and Validation, Loadig Image Data.

Day 9 | Transfer Learning

Learned how to use pre-trained networks to solved challenging problems in computer vision. Used densenet121 model to train a model to find the difference between a cat and a dog. Ipython Notebook of Transfer Learning.

Day 10 | MLP Classification

Build a model to recognise numbers using MLP Classification. IPython Notebook of MLP

Day 11 | MLP with validation

Build MLP with validation, learned about local connectivity, filters and the convolutional layer. IPython Notebook of MLP with validation

Day 12 | Sobel Operator and Convolution Layer

Created some sobel filters and learned the importance of filters. IPython Notebook of Sobel Filter

Day 13 | CNNs for Image Classification

Learned how CNN is used for image classification.

Day 14 | Build CNN model using SIFAR data

Build a 3 conovlution layers CNN model to predict 10 lasses, and learned how image augmentation helps in minimizing overfitting. IPython Notebook of CIFAR CNN, Augmentation

Day 15 | Style Transfer

Learned about the style transfer technique using CNN and how gram matrix style transfer works.

Day 16 | VGG19 Model

Build a style transfer model using VGG19 network. IPython notebook of Style Transfer

Day 17 | Recurrent Neural Networks

Learned about Recurrent Neural Networks from CS231n

Day 18 | Google Cloud On Board India

Attended the Google Cloud on board event where learned about the big data tools and how to implement real time analytics on Google Cloud. Attended the session on Auto ML and learned about the Vission API and NLP Api by google.

Day 19 | LSTM

Learned about the Architecture of LSTM.

Day 20 | GRU

Learned about GRU from Michael Guerzhoy's post

Day 21 | Time Series Prediction

Learned about time seris predictions. IPython notebook of Simple RNN

Day 22 | Characterwise RNN

Learned the characterwise RNN model.

Day 23 | Character-Level RNN

Building a RNN model to predict novel phrases.

Day 24 | Character-Level RNN

Completed the model. IPython notebook of Char RNN

Day 25-29 | Sentiment Prediction with RNN

Build a project on IMDB data.

Day 30-36 | Image Classifier Project

Working on the final Project.

Day 37-39 | Image Classifier Project

Work in progress

Day 40-49 | Pytorch Scholarship Finished

Finished the project, completed a case study on Time series analysis.