I'm a software developer with an interest in machine learning. Currently, I work for a media monitoring company where I help develop web and media processing applications. Our systems monitor over 200 radio and TV stations across the Caribbean.
I also hold an MSc. in Computer Science from The University of the West Indies, St. Augustine. For my final project I used convolutional neural networks to classify images of local hot peppers. A well-known interpretability technique called Grad-CAM was used to examine how the model made predictions.
Previously I worked as a programmer at a payroll software company. There, I mainly did backend programming for the company's web application. I have also worked as a full-time teaching assistant at the University of the West Indies, St. Augustine. At UWI I tutored introductory to advanced programming classes and conducted programming labs with undergraduate students.
My dream is to work on software that can improve people's lives. Projects where the end goal is not profit or a business objective, but to empower people to do things they could not do before.
"An Annotated Caribbean Hot Pepper Image Dataset"
Arai, K. (eds) Proceedings of the Future Technologies Conference (FTC) 2022, Volume 1. FTC 2022 2022. Lecture Notes in Networks and Systems, vol 559.
This paper involved collecting images of local hot pepper fruit with annotations of their physical features. It also showed a simple image classification task using CNNs and the created dataset. This dataset can be used for further work in developing Deep Learning applications for the regional hot pepper industry. It additionally serves as the starting ground for a Caribbean based hot-pepper ontology.
Below is a short listing of some of the things I've worked on in my own time...
An extract from my project showing how the transfer learning technique can be used with the VGG16 model with weights after being pre-trained on ImageNet.
Transfer Learning with VGG16 in Keras
An elementary example of elliptic curve cryptography (ECC) in Python. ECC is the foundation of block chain technology.
Elliptic Curve Cryptography
An extract from my project showing an implementation of a CNN interpretability technique known as Grad-CAM.
Grad-CAM Demo
A browser based controller respose curve demonstration. Written in pure JavaScript. Uses the standard JavaScript GamePad API
Controller Response Curve Demonstration
An extract from a project showing a use of K-Means clustering to group reviewers based on average product rating given. This sample use the scikit-learn libray in Python:
K-Means Clustering sample
A small project I did for fun when GameStop stock (GME) was going to the MOON! Uses a scheduled AWS Lambda to consume a stock API and stores the result in a DynamoDB table. An AWS API endpoint is used to access an AWS Lambda function to return an HTML page to the user. The content of the page determined by the price change in GME stock.
GME Price Tracker
A small project I did for fun. Using a data visualization library known as HighCharts.js to create a radial pokemon type chart.
Radial Type Chart
Check out my other repositories at GitHub...
GitHub