In this article we try regression and prediction using PyMC3 and a concept called Gaussian processes. And this famous german guy Gauss? How smart was he, really?
Making good predictions is difficult, predicting the future is even harder. But let's start with a simple curve fitting example using probabalistic programming and go from there.
The Birding Forecast estimates the likelihood of finding a bird species on a specific birding location. When during the year are the chances higher to find a specific species and what species are most likely to see right now?
Could data science help you in your birding efforts?
Perhaps, otherwise just check out the graphs to confirm what you already knew.
So let's dive into crunching the data from a birders perspective.
The Mystery Birdname Quiz creates completely fictitious birdnames using AI/Machine Learning. Mixed with random real names it makes a bird quiz application.
The LSTM-model used was trained in a Python environment and converted to TensorFlow JS to run in the web browser.
Visualization of the estimated reproduction rate (Rt) and Mobility. Also check out charts showing seasonal induced Virus Decay vs Rt, IFR, PCR-Testing volume, proportion of positive cases and Anti-Body levels.
Statgenic Data Science
E-mail: statgenic at knutas.com