Aakash Kaku's repositories
intermdiate_layer_matter_ssl
The official repository for "Intermediate Layers Matter in Momentum Contrastive Self Supervised Learning" paper.
knee-cartilage-segmentation
The project covers various deep learning models to automate the segmentation of knee cartilages using the diffusion weighted MRI
seq2seq_hrar
The official repository for "Sequence-to-sequence modeling for action identification at high temporal resolution"
Yelp-recommendation-engine
The project compares different recommendation algorithms for recommending restaurants to Yelp users.
bag-of-n-gram-model
Bag of n-gram classifcation model to perform sentiment analysis of IMDB reviews
SNLI
Natural Language Inference Task
Bank-Marketing-Project
Abstract—Data from a marketing campaign run by Banco de Portugal is examined. The campaign’s aim was to increase customers’ subscription rates to fixed-term deposit products, such as CDs. Using knowledge from the course, a number of machine learning algorithms are implemented to answer the question: How can banks successfully market these products in the most efficient way possible and with the highest possible rate of success? I. INTRODUCTION With the startling rise over the last few decades of media and technology which increases the amount of information we have at our fingertips (cell phones, television, Internet, etc.), humans are now more connected than ever. One result of this is that marketing campaigns are growing evermore pervasive in our daily lives. This glut of advertising has forced businesses to compete for the attention of a populace that has an ever growing amount of distractions. Thus raising the question: How can businesses successfully advertise their products in the most efficient way possible with the highest possible rate of success? We will answer this question in the context of banks advertising fixed term deposit products to their customers. Using data collected from a previous bank marketing campaign, a number of features centered around the clients, the campaign itself, and general market conditions will be explored. Based on this data, machine learning models will predict which clients will subscribe and what banks can do to increase the rate of subscription.
Data-InQuirer
Big Data Course Final Project
ELR
Official Implementation of Early-Learning Regularization Prevents Memorization of Noisy Labels
python_programming
python_programming
test
test repo
udvd
Unsupervised Deep Video Denoising, ICCV 2021