[3Xin2] Poster session 1

[3Xin2-37] Stock to Music: Transformation of Multivariate Stock Time Series into Music Data

〇Yuki Hiramatsu¹, Kei Nakagawa², Kaito Takano², Eita Nakamura³ (1.Mitsui Sumitomo Aioi Life Insurance Co., Ltd., 2.Nomura Asset Management Co., Ltd., 3.Kyoto University)

Keywords:music generation, finance, investment

We propose a method for transforming multivariate stock time series data into music. Typically, investors rely on visual information, such as stock charts, to make investment decisions based on stock time series. However, watching numerous stock data simultaneously and for long time is highly demanding and challenging. Therefore, we focus on representing essential information for investment decisions (trends and sudden changes) through sound information, or more precisely, music, which does not depend on visual interpretation. We expect that decision-making through musical transformation can reduce cognitive load and be accessible to a broader range of users. We conduct experiments to determine whether the essential information can be successfully acquired from the music generated by the proposed method.