Human In Command Machine Learning

Abstract

Machine Learning (ML) and Artificial Intelligence (AI) impact many aspects of human life, from recommending a significant other to assist the search for extraterrestrial life. The area develops rapidly and exiting unexplored design spaces are constantly laid bare. The focus in this work is one of these areas; ML systems where decisions concerning ML model training, usage and selection of target domain lay in the hands of domain experts. This work is then on ML systems that function as a tool that augments and/or enhance human capabilities. The approach presented is denoted Human In Command ML (HIC-ML) systems. To enquire into this research domain design experiments of varying fidelity were used. Two of these experiments focus on augmenting human capabilities and targets the domains commuting and sorting batteries. One experiment focuses on enhancing human capabilities by identifying similar hand-painted plates. The experiments are used as illustrative examples to explore settings where domain experts potentially can: independently train an ML model and in an iterative fashion, interact with it and interpret and understand its decisions. HIC-ML should be seen as a governance principle that focuses on adding value and meaning to users. In this work, concrete application areas are presented and discussed. To open up for designing ML-based products for the area an abstract model for HIC-ML is constructed and design guidelines are proposed. In addition, terminology and abstractions useful when designing for explicability are presented by imposing structure and rigidity derived from scientific explanations. Together, this opens up for a contextual shift in ML and makes new application areas probable, areas that naturally couples the usage of AI technology to human virtues and potentially, as a consequence, can result in a democratisation of the usage and knowledge concerning this powerful technology.

Restaurant Sales Prediction Using Machine Learning

“…Another effective mechanism that was utilized by (Holmberg and Halldén, 2018) in order to achieve forecasting or sales in restaurant based on the concepts of machine learning is considered as a one of the most appropriate techniques in the domain of restaurants. In the research study three different dataset were utilized for the purpose of experimentation based on three different cities.…”

NEURAL NETWORKS IN CONTEXT: CHALLENGES AND OPPORTUNITIES A Critical Inquiry into Prerequisites for User Trust in Decisions Promoted by Neural Networks

“…By focusing on these, from an average case perspective, state of the art ML models we aim at illuminating deficiencies that follow an objectivising stance. Our focus is based on a human in command relation to AI/ML and, consequently, a focus on subjective understanding (Holmberg, 2021b). XAI, to be useful, needs as a minimum to function as a trustworthy tool for our selected group of younger persons with IT-related education, if we are to expect them to be useful for other groups.…”

“…None of the reviewed papers includes a trained ML model that focuses on entity-data relations, relations that a neural network can learn well in a similar fashion as it can learn kind-data relations. Entity-related concepts follow the instance and can be related to ageing or wear and tear (e.g., 'scratches', 'splinters', or 'makings') and can be useful to, for example, identify objects and estimate ageing (Holmberg, 2021b).…”

“…Our focus is on classification tasks for images picturing mundane visual objects, a setting in which the strengths and weaknesses of the ML-system are less obscured compared to a more demanding domain. We thereby construct a setting in which a human with domain knowledge can be expected to assess reasons for a decision presented by the ML-system and therefore over time and usage build trust and knowledge concerning the system's strengths and weaknesses (Holmberg, 2021a).…”