Word Embeddings
Nov 2, 2019
Summary of word embeddings
word2vec glove fasttext import numpy as np
Mitigation of unintended bias in text classification
Aug 9, 2019
Kaggle comprtition organized by Jigsaw Conversation AI. It was second edition of the competion.
Bias Metrics BPSN SPBN Subgroup Combined tbc
Mod import numpy as np
A bottom-up recommender system for travel destinations
Aug 3, 2019
– Project Status: Active Objective To develop a web based platform that gives personalized recommendations for travel destinations and activities based on user preferences. The recommendations will be accompanied with a list of directly comparable features pertaining to the suggested destinations, such that the user(s) can delve deeper and emerge with a satisfied choice.
To this end, a bottom-up solution algorithm is adopted:
At the top level, a list of countries are considered Each country contains a list of travel destinations Each destination contains a list of Points Of Interest (POI) We will predict the scores of the POIs based on user profile.
...
➦
Neural Machine Translation using Encoder-Decoder Networks
Jul 12, 2019
– Project Status: Active In this project, I aim at building different sequence to sequence machine learning models for Spanish to English translation task using PyTorch. The Sequence to sequence or “Seq2Seq” is an end-to-end model comprising of two recurrent neural networks:
An Encoder: takes sentences in the source language as input and encodes them into finite dimensional context vectors;
A Decoder: uses the context vector as a seed from which the translated sentences are generated.
...
➦
National Data Science Challenge 2019 - Shopee
Apr 10, 2019
Shopee, an e-commerce platform headquartered in Singapore, organized the National Data Science Challenge during February-March 2019, highlighting Singapore’s emphasis on driving the digital economy and the use of big data. I actively participated in this competition to apply machine learning skills for a real world commercial application. It was an enriching experience, with a bonus that my team (NYRS) secured the third position in the competition leaderboard. This article summarizes steps followed to solve the problem.
...
➦
Google Compute Engine setup for data science projects
Feb 1, 2019
Google Compute Engine (GCE) is a component of Google Cloud Platform (GCP), which enables users to launch virtual machines (VMs) on demand. As per the project requirements, VMs can support the use of SSD and GPUs. In this article, I will summarize how to setup a GCE VM for solving data science related problems.
NB: this document is for my future reference.
Steps The process involves the following steps:
...
➦
A micromorphic computational homogenization framework for composite materials
Oct 10, 2018
This article summarizes my contributions in the field of computational mechanics during (first-part of) my graduate studies at National University of Singapore (NUS). My work resulted in development of a novel multi-scale computational homogenization framework for composite and auxetic structures. For this novel contribution, I achieved a Best Student Paper Award during the 2017 American Society of Civil Engineers EMI Conference in San Diego. To date, this work have been published in two top journals in mechanics, namely Journal of the Mechanics and Physics of Solids (JMPS) and Computer Methods in Applied Mechanics and Engineering (CMAME).
...
➦
