Tags » R-project

Live model: Indiana Dem primary

Find my actual projections HERE.

7:49 pm CT:

Sanders’ lead is at 6 points with 57% votes in. The live model stands at Sanders +5.2. 256 more words

Machine Learning

Final projections: Indiana democratic primary

Get excited!

I have a real-time model in place now, which will keep on predicting and updating statewide margins as new voting data keep coming in from Indiana today. 236 more words

Statistics

yorkr is generic!

The features and functionality in my yorkr package is now complete. My R package yorkr, is totally generic, which means that the R package  can be used for all ODI, T20 matches. 352 more words

R Language

Indiana democratic primary predictions and a 'polls-plus' model

TL; DR: It is going to be a close contest, but Bernie Sanders is probably going to squeak through.

Since Indiana holds an open primary, and there is a high amount of independent voters in the state (19.6% according to a… 374 more words

Machine Learning

April 26 Democratic primary projections

It’s primary time again! My last week’s projections for the New York primary fell flat on its face and polls won the day, mostly because the model built on demographics and facebook data ended up overestimating upstate vote margins for Bernie Sanders. 239 more words

Machine Learning

New York democratic primary: final projection and county analysis

I made some minor changes to my previous model and the final numbers for NY come out to be slightly different. I have also calculated standard errors for the projection using the jackknife-based methodology I had used in my… 549 more words

Machine Learning

Democratic primary projections at county-level leveraging social media data

New York projections

Based on my new county-level model (details below), I am calling a 5.8-point win for Hillary Clinton in the New York Democratic primary.  639 more words

Machine Learning