Anzeige
Erschienen in:
2021 | OriginalPaper | Buchkapitel
Move Ahead with Mahalanobis Distance - Pattern Approach
verfasst von : Shuichi Fukuda
Erschienen in: Human Interaction, Emerging Technologies and Future Applications III
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Abstract
Human movement is composed of external motion and internal motor. Motion is controlled rationally with the fixed goal. So, tactics is important. But we, humans, coordinate our body parts instinctively to adapt to the environments and situations. Thus, coordination, i.e., strategy is important. Thus, to achieve our goal, we need to evaluate our performance and improve trial after trial. But quantitative performance indicator which support our instinct is not available. This paper proposes to utilize non-Euclidean Mahalanobis Distance (MD). Then, we do not have to worry about orthonormality and units, which Euclidean distance requires. It enables us to increase dimensionality easily and processes it in a very short time. Mahalanobis Distance-Pattern Approach proposed here combines MD and patterns. So, it enables a holistic and quantitative performance evaluation. Therefore, robots that work with warm consideration to our individual, personal way of movements can be developed beyond those just working for labor.