This Pede took up the mantle and ran with it... so much more than anyone here,,,
https://twitter.com/jessiprincey/status/1329525675666051073
This Woman got the Lawyer of the PRESIDENT to retweet what I've been posting here for days...
Jessi I hope I can meet you one day
great work
Congratulations and thank you. You’ve been promoted to General in Trumps Army!
Left Tenant Donkey Balls!!! spread this info
Roger that!
Awesome, bud. Great work and great to be acknowledged for it by the prez. Carry on!
Eh... I just read through the SMOTEboosting stuff and her inferences from that. It's wrong, atleast on that front. I read the paper, and it's an algorithm that builds on the popular boosting algorithm class. In the paper they compare their method to adaboost which is an algorithm I've personally implemented and worked with. Maybe this guy is related in a devious way, but his SMOTE stuff is almost certainly not directly related.
No one said SMOTE was it. Only that someone who wrote the book on SMOTE would be completely capable of doing it. Now, follow the links to his research and read the 10 papers he wrote on slowly predictably manipulating data. #DontBeLikeMSM
But all smote is related to is classification problems wrt to having imbalanced datasets where some classes in infrequently appear. I'm all in, just offering my thoughts
SMOTE is discussed simply to prove his work is based on topics that would be needed to programmatically alter votes as they are tabulated.
its almost beyond circumstantial is what I'm saying. I wish I had a proper analogy ready that would illustrate the difference in the skills and knowledge to do this work and the type of work in "altering votes programmatically as they are tabulated."
Maybe I'm wrong, but my 8 years in software development in which I spent every day building new systems for time series analysis and algorithmic trading, I'm struggling to connect the dots you're eagerly doing. But I do trust Powell, so there's that contradictory information I guess.