temcavanagh Goto Github PK

aws_system_design

bankingfrauddetection

Fraud detection using the Banksim dataset.

crypto-prediction

A predictive model for forecasting the hourly Bitcoin price movements using sentiment analysis drawn from extracted from Twitter data and tree based algorithms. The resulting predictive model utilises an XGBoost gradient boosted random forest model which demonstrates a 63.16% accuracy in predicting the hourly price movements of Bitcoin.

database-design-management

Projects and coursework as part of COMP6140 Database Design & Information Management

lpatternidentification

A python package for identifying unique class feature patterns from binary class datasets.

node-app

opencv

OpenCV setup

relativeneighborhoodgraph

A python package for computing the Relative Neighborhood Graph from a given distance matrix.

rubbish-classification

A PyTorch computer vision model for the detection and classification of rubbish in the environment.

seng6110-oop

Object Oriented Programming with Java

skin-cancer-detection

Implementing and comparing ResNet50 and MobileNetV2 transfer learning models using the MNIST:HAM10000 image dataset. Resulting classification accuracy of ~90%.

spark_etl_pipeline

Building an ETL pipeline using Apache Spark (spark-sql) for use in predictive modelling.

two-spiral

Comparison of ANN and SVM model performance on two-spiral classification tasks

wilt-r-analysis

A Comparative Assessment of Machine Learning Classification Methods For The Detection of Wilt Disease in Pine and Oak Trees in R