Repository for all my D3 Data-Viz Projects
View the projects here: https://ksun0.github.io/data-visualizations/. Special thanks to HTML5 UP! for providing the beautiful landing page.
Data processing and models I used are available in the Jupyter Notebooks of each project folder!
- Solar Map of Massachusetts
- View the degrees of solar radiation (GHI) at various coordinates in Massachusetts.
- Move the time slider to see GHI over time during the year of 2015!
- This visualization can be used to predict the best location of solar panels and perhaps outline some geographic boundaries of Massachusetts.
- Categorizing Science Articles with Natural Language Processing (NLP)
- This visualization shows how k-means clustering classifies various science articles.
- The real category scraped from the actual website is shows as well to compare.
- Take a look at the Jupyter notebook to see more about the text precessing and the ML model training that was involved.
- Twitter Sentiment Analysis to Predict Airline Reputation
- In this visualization, I use classic twitter sentiment analysis (a naive bayes model trained on airline tweet training data) to plot negative tweets over time.
- Large spikes in negative tweets correlates to a disastrous event that ruined an airline's reputation, for example a forced removal from the aircraft.