Machine learning applications in architecture: Opportunities, challenges and foresights for the future

Machine learning, one of the sub-branches of artificial intelligence, is defined as a system that can learn by analyzing data using algorithms, can imitate human intelligence and can improve itself. It is possible to store large amount of new data and the data formed by processing these data using machine learning. Although, machine learning techniques are used in many fields due to these features, especially in the engineering and production sectors, they are not yet widely used in the field of architecture because of scarcity of research and insufficient studies. The aim of this thesis is to highlight the benefits provided by machine learning in architecture and to identify the problems and risks that may be faced during design production and construction process. In this context, first the machine learning techniques are observed and their use in various disciplines are discussed. Then, examples carried out until today on machine learning in architecture are presented. In the thesis, the benefits that can be gained, problems that can be encountered and the risks that may rise during the use of machine learning are identified, focusing on how and for what purposes machine learning techniques are used in the field of architecture. It is concluded that the use of machine learning techniques in the field of architecture can be developed and used as an efficient tool in design, production and construction processes, depending on the findings of this research study.