The knowledge base of machine learning, across data analytics teams in a matrix organization. : An exploratory case study on machine learning

• Main Author: Johansson, Josefin
• Format: Others
• Language: English
• Published: Karlstads universitet, Handelshögskolan 2017
• Subjects: Other Engineering and Technologies not elsewhere specified; Övrig annan teknik
• Online Access: http://urn.kb.se/resolve?urn=urn:nbn:se:kau:diva-66283

Machine learning is a field within the broader concept of artificial intelligence and addresses the questions of how to build systems which learn from experience. The field is one of the oldest disciplines in computer science but has had many recent advancements due to the large amounts of data being generated. Today, machine learning together with artificial intelligence is seen as the two most rapidly growing fields within computer science. The purpose of this thesis is to explore and identify the current knowledge base of machine learning across data analytics teams, within the matrix organization Wise Inc.. This study has been performed using an exploratory case study method, based on the embedded units within the matrix organization. In this research, the units represent thirteen different cross-functional teams existing within the Wise Inc. organization. All thirteen teams are data analytics teams and performing a variety of different analytics depending on the team’s individual purpose. The analysis of embedded units has been performed within the units, but also across units. Using data collected through a qualitative questionnaire and interview, the knowledge base of machine learning could be explored and identified. Analysing the collected data, it was showed that the knowledge base across the data analytics teams in Wise Inc. is currently relatively low. Two key teams have been identified to have a very high level of knowledge. The knowledge base was examined based on participants theoretical and practical knowledge when it comes to machine learning. The aspect of machine learning usage and experience was included in the analysis and appeared to show a weak positive correlation to the overall knowledge. However, the statistical significance could not be determined. The empirical study also indicates that across teams, the level of knowledge is slightly higher than the level of experience. As a positive result, most participants appear to have a good theoretical understanding of machine learning in relation to artificial intelligence, which normally is one of the most common miss-interpretations. Even though the overall knowledge base is low, there are a few key people which stand out with a high knowledge base amongst teams. Observing the team as a whole the knowledge base is medium, but when looking at the individuals within the team there are a few key members with high expertise. These people are not working within the two teams identified with a high machine learning knowledge base but are part of other analytics teams. These people are important to identify as they can contribute with great value to the Wise Inc. organization.