GitHub dataset

MovieLens Dataset to use with In-Memory Analytics benchmark of CloudSuite

SIFT based image classification on the CIFAR-10 dataset.

Different Adversarial attack methods implemented in PyTorch on CIFAR-10 Dataset. This repository contains Adversarial Attacks on CIFAR-10 dataset implemented in Pytorch: It will include more Adversarial Attacks and Defenses Technique in future as well.

Support material for my OSDC London 2016 talk on Dataset Shift in Machine Learning. Supporting material for my OSDC London'16 talk on Dataset Shift in Machine Learning. Interactive plot to illustrate the phenomenon of covariate shift.That is, p(x,y) differs from training to testing phase. The interactive plot allows youto specify p(x) via a probability tabels while p(y|x) remains fixed.You can see that underspecified discriminative models are not immune to covariate shift.

Processing annotations for CULane dataset. It could:(a) generate per-pixel labels from original annotation files.(b) generate list files for training.

Code for the collection and analysis of the MTNT dataset. This repo contains the code for the EMNLP 2018 paper MTNT: A Testbed for Machine Translation of Noisy Text. It will allow you to reproduce the collection process as well as the MT experiments. You can access the data here.

Image classification using a variety of methods on MNIST dataset. Image classification using a variety of methods on MNIST dataset, including SVM, KNN, Decision Tree and CNN.

Use Faster-RCNN + FPN to detect objects on aerial image dataset(DOTA) This algorithm adds FPN(Feature Pyramid Network) into Faster-RCNN.I have tested on DOTA Dataset,which is an aerial image dataset provided by Wuhan Univeristy(China).