In this competition funded by The Learning Agency Lab and Vanderbilt University, competitors must use written student summaries to predict both wording and content scores. With models that can generalize well across prompts and grade levels (3-12), educators can leverage these predictions to augment their grading process and improve learning outcomes.

In this DrivenData competition, participants must create models to predict H1N1 and Seasonal Flu vaccinations. With predictive insights, public heath agencies can better understand characteristics that influence vaccine uptake and strategically improve response rates in future events.

In this Zindi competition, early learning assessments of South African children are used to predict early learning scores and identify elements of successful early learning programmes. With these models and their predictive insights, South Africa can greatly improve EL programmes and the developmental outcomes of generations to come.

Through the utilization of the Selenium library, these python scripts enable the user to interact with data sources such as NSERC's research awards database and Shanghai Ranking's online tables, and ultimately extract detailed data that is otherwise not exportable.

Using Tableau, this collection of data visualizations allows you to walk through the history of NHL trades. From analyzing shifts in trade assets to highlighting general manager track records - there are many ways to explore the data and enjoy some special moments in NHL history.

A curation of Javascript and Python data visualizations that I've created over time. Anything from a Hexagon Map of Canadian province stats to Raincloud charts with NHL player stats can be found here, but not too much more to be honest.