Published Apr 20, 2021

SDS 463: Time Series Analysis — with Matt Dancho

Matt Dancho, founder of Business Science, delves into the transformative journey of his company from consulting to education, exploring innovative tools like Timetk and Modeltime for time series analysis and the integration of R Shiny for building impactful web applications, while highlighting deep learning's role in forecasting with models like Nbeats and DeepAR in diverse data science projects.
Episode Highlights
Super Data Science: ML & AI Podcast with Jon Krohn logo

Popular Clips

Episode Highlights

  • Shiny Apps

    Matt explains how R Shiny enables the creation of data-centric web applications with high customizability. He shares his experience in developing courses and consulting on R Shiny, emphasizing its evolution and power in integrating time series analysis into applications 1. One notable example is his Nostradamus app, which automates forecasting using time series packages like Modeltime and Timetk 2.

    Shiny, the ecosystem has evolved so much, it's insane, and it's just so powerful.

    ---

    Jon highlights the ease of building and deploying these applications, making R Shiny a valuable tool for data scientists.

       

    RStudio Connect

    RStudio Connect plays a crucial role in the deployment and management of R Shiny applications. Matt praises the RStudio IDE for its user-friendly interface and extensive support for both R and Python 3. He shares a personal anecdote about meeting Hadley Wickham, the developer of essential R packages like Dplyr and Ggplot, highlighting Wickham's influence on the R community 4.

    Without them, R probably wouldn't exist. It would be all Python, to be honest.

    ---

    Jon and Matt agree on the importance of RStudio in advancing data science practices.

Related Episodes