Declan-Curran1
A hard-working computer science/economics student with a strong basis in statistics/math
Brisbane
Pinned Repositories
html_feedback
Supports a student project developing a UI for feedback on arXiv articles rendered as html.
Insider-Trading-Task
Completed a project in late 2020 where the effects of insider trading on stock price were analysed. Analysis was done in R on company GABC
Natural-Language-Processing-From-Tweets
My submission for the "Determining Natural Disasters from Tweets" competition on Kaggle. Used a ridge regression in python to analyse text from tweets in order to determine if the tweet was referring to a natural disaster or not. Was able to predict with 65% accuracy.
Real-Time-Bushfire-Prediction
Takes high resolution multispectral sentinel-2 satellite imagery. Applies k-means clustering to group pixels into seven groups based on the 12 bands available with Sentinel 2. Uses time-series (data collected every 3 days between both Sentinel 2 satellites) on the Sydney/blue mountains area to predict whether clusters were predictors for fire
Titanic
My submission to the "Titanic" competition held on Kaggle. Created a random forest model in R to predict whether a given passenger would survive the sinking of the titanic
WeatherLearn
Implementation of the PyTorch version of the Weather Deep Learning Model Zoo.
ClimateDiffuse
Diffusion for climate downscaling
a-PyTorch-Tutorial-to-Super-Resolution
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network | a PyTorch Tutorial to Super-Resolution
Declan-Curran1's Repositories
Declan-Curran1/Insider-Trading-Task
Completed a project in late 2020 where the effects of insider trading on stock price were analysed. Analysis was done in R on company GABC
Declan-Curran1/Natural-Language-Processing-From-Tweets
My submission for the "Determining Natural Disasters from Tweets" competition on Kaggle. Used a ridge regression in python to analyse text from tweets in order to determine if the tweet was referring to a natural disaster or not. Was able to predict with 65% accuracy.
Declan-Curran1/Real-Time-Bushfire-Prediction
Takes high resolution multispectral sentinel-2 satellite imagery. Applies k-means clustering to group pixels into seven groups based on the 12 bands available with Sentinel 2. Uses time-series (data collected every 3 days between both Sentinel 2 satellites) on the Sydney/blue mountains area to predict whether clusters were predictors for fire
Declan-Curran1/Titanic
My submission to the "Titanic" competition held on Kaggle. Created a random forest model in R to predict whether a given passenger would survive the sinking of the titanic