I'm a data scientist and machine learning practitioner.
I have a Bsc in Mechanical Engineering, but the variety of problems
I'm interested in applying ML and Data Science tools to is quite broad.
Below are a few of the projects I've been involved in.
Nursing Home Brief IoT Monitoring: Omdena, DriQ Health
This project was done as a volunteer ML Engineer for Omdena,
a datascience volunteering platform. The client was a startup in the
healthcare monitoring space called DriQ Health. Their mission was to develop
a system that would allow nursing home staff to determine if seniors
were wearing wet diapers without having to check them manually.
Seniors left to wear unchanged briefs for extended periods are at
risk of skin breakdown, and dangerous urinary tract infections.
I was involved in the dataset creation and labeling pipeline, and model evaluation.
Note: the certificate title is for "Lead ML Engineer" due to a title upgrade at the end of the project.
This project was done as a volunteer Junior ML Engineer for Omdena,
a datascience volunteering platform. The client was a startup in
the regenerative agriculture space named Azolla Projects. Their
mission was to help farmers transition to regenerative agriculture
methods, since these methods can decrease a farm's carbon emission,
thereby generating carbon credits that can be sold in carbon
markets. We were tasked with creating tools to specifically
visualize and predict the effect of certain methods on a
given plot's soil organic carbon (SOC). I was involved in
data acquisition and research, and was one of the main
members of the modeling team.
Machine Reliability Prediction Using Sound (Unfinished)
This is a model that, given a sound clip from an operating machine,
can classify that operation as either normal, or abnormal.
It is trained using data from the MIMII machine sound dataset,
which contains sounds from several classes of machines.
Solar Power Prediction (AMII ML Technician Course)
This is the final presentation created by my team in the AMII ML
Technician Course. Our team was given a dataset, and over the
course we spent time scoping out, and evaluating an business
case for an ML solution, then analyzing the data and building
and evaluating an appropriate model.
A web application where the user can compete against a convolutional
neural network to see who can more accurately identify whether
presented pictures of possibly cancerous skin lesions are malignant
or benign. This project was done with the Undergraduate AI Society
at the University of Alberta, for the AI in Medicine Student
Society Symposium.
Note: the currently linked
version of the webapp uses pre-computed predictions.