Statistics and Data Analysis

Many applications in industry, medicine, biology, finance and marketing rely on measurements but it is not always the case that the data is exploited and utilized to its full potential. FCC provides knowledge for such utilization using state-of-the-art tools and methods. For the customer this facilitates the understanding of the processes and inherent variation involved in diverse applications, whether data comes from fatigue loads on agricultural equipment, tracking and measurements of protein levels on hundreds of individual cells over several hours, or robust design methodology for assembled complex products.

In modern applications where data set sets are huge, advanced mathematical and computational tools facilitate detection of patterns and relationships.

In many modern applications where data is gathered for example from internet feeds or from process measurements the dimensionality of the data set easily becomes overwhelmingly large. In such applications it is hard or impossible for a human being to grasp even a small portion of the data, in order for example for detecting relationships and patterns and for decision support. Knowledge discovery and data mining in data like this require advanced knowledge both in mathematical statistics as well as computational and algorithmic tools. It involves concepts and methodology from supervised or unsupervised statistical and machine learning, including various classification and clustering techniques, such as neural networks, support vector machines, classification trees and boosting. At FCC we have this competence and combine standard and in-house developed software to conduct projects in this area.

© 2014-2017  Fraunhofer-Chalmers Centre