-
Book Overview & Buying
-
Table Of Contents
How to Prove Anything
By :
Due to the meticulously cleaned data, black box time analysis tools were easily applied and evaluated against Tiffany’s historical data. With more than two years of data, these forecasting models could be cross-validated for a historical first in our ten-year relationship, far beating the overfitted multivariate approach, which caused the end of my relationship with my high school sweetheart a year into college [5]. Tiffany’s data was modeled in this paper using a seven-day moving average, sextuple exponential smoothing, Autoregressive Moving Average (ARMA), and one overly complicated machine learning black box.
The most simple model applied to Tiffany’s mood swing data was a seven-day moving average. While this extremely rudimentary approach may not have been the best for implementing higher dimensionality predictors, it created less-noisy forecasts compared to the more complex alternatives.
While her data appeared...
Change the font size
Change margin width
Change background colour