Metrological machine learning (2ML)


convergence: models are data

models are perfectly observable -> no black boxes

complex (not easy to understand) but completely transparent (unless behind api)

data optimality: research questions, cost and performance, impact (applications)

metrology, measurements as the answer: we can observe -> we need to measure

impact implications in research and deploy:

answer what is optimal question

answer under what conditions questions

solve tracking atc


conclusions and implications

data internet

logic/theory: map distributed insights onto the integrated data transformation lifecycle

data engines

insight/discipline integrator to answer what is questions

detailed rundown and connections and explanations