Predictive Modelling: From Naive Bayes to Deep Learning
David Meintrup, Professor of Mathematics, Statistics and Operations Research at Ingolstadt University of Applied Sciences
Empirically modelling data is mission-critical for achieving success in today’s competitive environment. All kinds of engineering and technical challenges can now be solved (and even anticipated) by leveraging the data that’s generated by routine operations.
This session focuses on making predictions using classification and deep learning techniques.
‘Classification’ tries to assign objects to classes using the measurements made. If the classes are known in advance this is called supervised learning; if not, it is unsupervised learning.
‘Deep learning’ is a recent branch of machine learning and has been responsible for numerous successful applications, from image classification to autonomous driving. It involves algorithms that are able to learn and improve in an autonomous way.