anishacharya Goto Github PK

anishacharya

Try out git README

anishacharya.github.io

bandits-online-learning

Simple Implementations of Bandit Algorithms in python

bgmd-aistats-2022

Geometric median (GM) is a classical method in statistics for achieving a robust estimation of the uncorrupted data; under gross corruption, it achieves the optimal breakdown point of 0.5. However, its computational complexity makes it infeasible for robustifying stochastic gradient descent (SGD) for high-dimensional optimization problems. In this paper, we show that by applying Gm to only a judiciously chosen block of coordinates at a time and using a memory mechanism, one can retain the breakdown point of 0.5 for smooth non-convex problems, with non-asymptotic convergence rates comparable to the SGD with GM.

contrastive-positive-unlabled-pu-learning

cracking-coding-interviews

Contains Basic Utility Functions and Common DS and Algo Questions

cs273a-uci-fall-2013

My First Graduate Machine Learning Course- 2013 Fall :)

cvd-19-experiments

dbms-from-scratch

Implementation of a DBMS working system ground Up

ddp_training

deepcluster-contrastive

Deep Clustering for Unsupervised Learning of Visual Features

delicoco-ieee-transactions

In compressed decentralized optimization settings, there are benefits to having multiple gossip steps between subsequent gradient iterations, even when the cost of doing so is appropriately accounted for e.g. by means of reducing the precision of compressed information.

dlrm

An implementation of a deep learning recommendation model (DLRM)

double-descent

We investigate double descent more deeply and try to precisely characterize the phenomenon under different settings. Specifically, we focus on the impact of label noise and regularization on double descent. None of the existing works consider these aspects in detail and we hypothesize that these play an integral role in double descent.

expectation-maximization

Package in Matlab for generating Synthatic Data using GMM and EM Clustering on that

federated-learning-in-pytorch

Handy PyTorch implementation of Federated Learning (for your painless research)

image-segmentation-fractional-filters

Software Related to CVPR 2015 Paper on Image Segmentation Using TRW Belief Propagation based Learning

learning-with-audio-data

My Functions to easily create music dataset from raw signals, run learning algorithms and classify genres

lightly

A python library for self-supervised learning on images.

llm-decoding

nlp-cs388-ut

Common Libraries developed in "PyTorch" for different NLP tasks. Sentiment Analysis, NER, LSTM-CRF, CRF, Semantic Parsing

online-embedding-compression-aaai-2019

Deep learning models have become state of the art for natural language processing (NLP) tasks, however deploying these models in production system poses significant memory constraints. Existing compression methods are either lossy or introduce significant latency. We propose a compression method that leverages low rank matrix factorization during training, to compress the word embedding layer which represents the size bottleneck for most NLP models. Our models are trained, compressed and then further re-trained on the downstream task to recover accuracy while maintaining the reduced size. Empirically, we show that the proposed method can achieve 90% compression with minimal impact in accuracy for sentence classification tasks, and outperforms alternative methods like fixed-point quantization or offline word embedding compression. We also analyze the inference time and storage space for our method through FLOP calculations, showing that we can compress DNN models by a configurable ratio and regain accuracy loss without introducing additional latency compared to fixed point quantization. Finally, we introduce a novel learning rate schedule, the Cyclically Annealed Learning Rate (CALR), which we empirically demonstrate to outperform other popular adaptive learning rate algorithms on a sentence classification benchmark.

optimization-mavericks

This repository provides a unified framework to perform Optimization experiments across Stochastic, Mini-Batch, Decentralized and Federated Setting.

robust-diffusion

search-engine-from-scratch

implementation of a search engine. Implements a complete search engine for the ICS domain.It also develops a[web] interface that provides the user with a text box to enter queries and returns relevant results

snclr

[ICLR 2023] Soft Neighbors are Positive Supporters in Contrastive Visual Representation Learning

vpu

A PyTorch implementation of the Variational approach for PU learning