Capstone Project – Native Language Classification by accent error detection. This project uses a scraped and self constructed dataset, run through both a neural network and several other machine learning classification models to predict native languages based on specific phonetic error types and locations.
Other Projects Predicting house prices with the Ames housing dataset. Finding the right parameters to achieve a successful kickstarter campaign. Hypothesis testing with SAT score and Drug Usage U.S, census data.
Data Job Postings Analysis Project Scraping data from job boards, analysing the language using Natural Language processing and an unsupervised machine learning model, LDA, found and visualised, with word clouds, words that predict high and low salary positions as well as Data Scientist positions vs. other Data jobs.
Robyn Hillman-Harrigan 
You must be logged in to post a comment.