Name: Asmita Poddar
Type: User
Company: University of Oxford
Bio: Graduate Student in Computer Science at University of Oxford. Machine Learning Researcher - AI for Healthcare.
Twitter: asmita_poddar
Location: Oxford, United Kingdom
Blog: https://sites.google.com/view/asmitapoddar
Asmita Poddar's Projects
Autoencoder-ATLAS: codes for implementation of autoencoder-decoder network to compress the data from the ATLAS experiments from 4 to 3 variables, and reconstruct the compressed data.
Creating package for simulation and classification of multi-dimensional spectroscopic data taken from Sentinel-2 satellite covering the area of France
Clinical Named Entity Recognition for EHR
Pytorch implementation of "Dynamic Coattention Networks For Question Answering"
Deep learning for identifying important motifs in DNA sequences. Exploration of the structure of the human genome by training neural network architectures (CNN, LSTM, Attention).
Festival Management
Sentiment Analysis with Ensemble
Data Analysis of Electricity Production and Consumption data of Romania over 9 years as a part of the EEML 2019 Data Analysis Challenge for Artificial Intelligence for Social Good
Named-Entity-Recognition-with-Bidirectional-LSTM-CNNs
#nowplaying-RS: Music Recommendation using Factorization Machines
Repository for Programming Assignment 2 for R Programming on Coursera
Implementation of a Dynamic Coattention Network proposed by Xiong et al.(2017) for Question Answering, learning to find answers spans in a document, given a question, using the Stanford Question Answering Dataset (SQuAD2.0).
Coded in C++ using OpenCV to achieve colour space transformation from RGB to YCoCg in order to increase embedding capacity, and then watermark extraction, so that both watermark and cover medium remain unchanged.
Some research experiments
An Open Source Machine Learning Framework for Everyone
Deblurring a text image using Convolutional Neural Networks
Time Series Analsyis of Wireless Network Parameters using ARIMA model
Crawler to retrieve and visualize user profiles based on gathered statistics
Takes feeds from Twitter into R and the sentiment of the tweets is analysed and classified into positive, negative and neutral tweets.
Estimating uncertainty of neural networks for automated screening of Diabetic Retinopathy using the PyTorch framework. Generated visual explanation of the deep learning system to convey the pixels in the image that influences its decision Integrated Gradient method.