Browsing by Author "Mahanama, Thilini V."

Now showing 1 - 3 of 3

Assessment of Human Emotional Responses to AI–Composed Music: A Systematic Literature Review
(Institute of Electrical and Electronics Engineers (IEEE), 2024) Fernando, Poorna; Mahanama, Thilini V.; Wickramasinghe, Manya
In the world of musical creation, the integration of artificial intelligence (AI) represents a significant paradigm shift in emotional engagement. This research investigates the human emotional responses evoked when listening to AI-composed music. Focused on figuring out the emotional impact of AI-composed music, the study explores the complex relationship between human emotional experiences and compositions crafted by AI algorithms. Through a comprehensive literature review, this paper examines existing methodologies, insights, and gaps in understanding the emotional dimensions of AI-composed music. Major findings reveal that while AI software like artificial intelligence virtual artist (AIVA) shows it can help explore emotional authenticity, ongoing doubt and preference for music made by humans highlight the need for more research. Attitudes of both listeners and music professionals toward AI-composed music are characterized by skepticism and negative perceptions, emphasizing the urgency to address reservations and investigate the unique emotional qualities of AI-composed music. Furthermore, the complex nature of music emotion recognition, influenced by factors such as music genre, cultural perspective, and age group, complicates understanding emotional responses to both human-created and AI-composed music. The paper supports the development of analytical methods, particularly through machine learning and deep learning approaches, to enhance understanding of the complexities of emotional responses and improve AI music composition. A human- experience-centered framework is proposed to address subjectivity in assessing emotional responses to music. This research aims to understand the details of emotional responses and find out if AI-composed music can really evoke emotions comparable to human-created compositions.
The Financial Market of Indices of Socioeconomic Well-Being
(2024) Mahanama, Thilini V.; Shirvani, Abootaleb; Rachev, Svetlozar; Fabozzi, Frank J.
This study discusses how financial economic theory and its quantitative tools can be applied to create socioeconomic indices and develop a financial market for the so-called “socioeconomic well-being indices”. In this study, we quantify socioeconomic well-being by assigning a dollar value to the well-being factors of selected countries; this is analogous to how the Dow 30 encapsulates the financial health of the US market. While environmental, social, and governance (ESG) financial markets address socioeconomic issues, our focus is broader, encompassing national citizens’ wellbeing. The dollar-denominated socioeconomic indices for each country can be viewed as financial assets that can serve as risky assets for constructing a global index, which, in turn, serves as a “market of well-being socioeconomic index”. This novel global index of well-being, paralleling the Dow Jones Industrial Average (DJIA), provides a comprehensive representation of the world’s socioeconomic status. Through advanced financial econometrics and dynamic asset pricing methodologies, we evaluate the potential for significant downturns in both the socioeconomic well-being indices of individual countries and the aggregate global index. This innovative approach allows us to engineer financial instruments akin to portfolio insurance, such as index puts, designed to hedge against these downturn risks. Our findings propose a financial market model for well-being indices, encouraging the financial industry to adopt and trade these indices as mechanisms to manage and hedge against downturn risks in well-being.
Stream Count Predictive Analysis for Upcoming Songs on Spotify using Machine Learning: A Systematic Literature Review
(Institute of Electrical and Electronics Engineers (IEEE), 2024) Ranidu, M.G. Yoshitha; Mahanama, Thilini V.; Wijenarayana, Sankini
In the era of evolving music consumption, this systematic literature review researches the realm of predictive analytics for music streaming, specifically targeting Spotify's stream count prediction in Sri Lanka through machine learning methodologies. With streaming platforms shaping the music industry landscape, accurately predicting song popularity becomes essential for artists, producers, and industry stakeholders. This review analyzes global studies on machine learning's application in forecasting stream counts while defining their methodologies and outcomes. It intricately examines diverse machine-learning methodologies employed in prior research endeavors. Ranging from regression models and ensemble techniques to deep learning architectures, the spectrum of methodologies used in forecasting stream counts on music streaming platforms is elucidated. Noteworthy techniques such as support vector machines (SVM), random forests, and recurrent neural networks (RNNs) have demonstrated efficacy in capturing intricate patterns within music data for predictive analysis. Our paper highlights the significance of feature engineering and selection methods, underscoring their pivotal role in enhancing the accuracy of predictive models. Through this comprehensive study, this review aims to expose specific gaps in stream count prediction models tailored to Sri Lanka's varied music preferences and consumption habits. By illuminating these gaps, it aspires to stimulate future research endeavors focused on refining predictive models, ultimately empowering the Sri Lankan music industry with more insights for better strategic decision-making.