Further enhancement of planned random algorithm applied in music shuffling

By: Carpio, Bea Mikaela N.; Francisco, Ricky Jr. O.; Gurimbao, Lance Angelo G
Language: English Publisher: . . c2025Description: Undergraduate Thesis: (Bachelor of Science in Computer Science) - Pamantasan hg Lungsod ng Maynila, 2025Content type: text Media type: unmediated Carrier type: volumeGenre/Form: academic writingDDC classification: . LOC classification: QA76.9 A43 C37 2025
Contents:
ABSTRACT: The Enhanced Planned Random Algorithm often encounters difficulties with data adaptability, precise similarity detection, and ensuring the even distribution of items. These challenges arise from its dependence on static attributes, exact attribute matching requirements, and an inherent inability to consistently space out similar items, which is particularly critical in applications demanding variety and a nuanced understanding of item characteristics. This research focused on overcoming these fundamental limitations by developing significant enhancements for the Enhanced Planned Random Algorithm by Jumaquin and Licudo (2024). The improved methodology incorporates dynamic attribute processing using a variance threshold method, enabling the automatic identification and prioritization of the most relevant item features, thereby moving beyond predefined attribute sets. To address similarity detection, K-Prototype clustering, complemented by Gower’s Distance, is integrated to facilitate a more sophisticated evaluation of item similarity across mixed numerical and categorical data types, allowing for the recognition of partial attribute matches. Furthermore, to ensure a more consistent and equitable distribution of similar items, the Martin Fiedler Algorithm is applied post-shuffling, promoting better arrangement and mitigating undesirable clumping. Comparative testing against the baseline FEPRA demonstrated marked improvements: the enhanced system dynamically selected a greater number of attributes (11-14 compared to FEPRA’s fixed five), achieved a significant increase in average consecutive Gower’s distance (from 3.78% to 7.00%), indicating superior item differentiation, and resulted in a substantially more uniform distribution of similar items, with reduction in the Standard Deviation of Gaps by as much as 86% to 91%. These synergistic advancements culminate in a more robust, adaptive, and perceptually satisfying item sequencing system, enhancing balance and variety for an improved listener experience.
Tags from this library: No tags from this library for this title. Log in to add tags.
    Average rating: 0.0 (0 votes)

ABSTRACT: The Enhanced Planned Random Algorithm often encounters difficulties with data adaptability, precise similarity detection, and ensuring the even distribution of items. These challenges arise from its dependence on static attributes, exact attribute matching requirements, and an inherent inability to consistently space out similar items, which is particularly critical in applications demanding variety and a nuanced understanding of item characteristics. This research focused on overcoming these fundamental limitations by developing significant enhancements for the Enhanced Planned Random Algorithm by Jumaquin and Licudo (2024). The improved methodology incorporates dynamic attribute processing using a variance threshold method, enabling the automatic identification and prioritization of the most relevant item features, thereby moving beyond predefined attribute sets. To address similarity detection, K-Prototype clustering, complemented by Gower’s Distance, is integrated to facilitate a more sophisticated evaluation of item similarity across mixed numerical and categorical data types, allowing for the recognition of partial attribute matches. Furthermore, to ensure a more consistent and equitable distribution of similar items, the Martin Fiedler Algorithm is applied post-shuffling, promoting better arrangement and mitigating undesirable clumping. Comparative testing against the baseline FEPRA demonstrated marked improvements: the enhanced system dynamically selected a greater number of attributes (11-14 compared to FEPRA’s fixed five), achieved a significant increase in average consecutive Gower’s distance (from 3.78% to 7.00%), indicating superior item differentiation, and resulted in a substantially more uniform distribution of similar items, with reduction in the Standard Deviation of Gaps by as much as 86% to 91%. These synergistic advancements culminate in a more robust, adaptive, and perceptually satisfying item sequencing system, enhancing balance and variety for an improved listener experience.

Filipiniana

There are no comments for this item.

to post a comment.

© Copyright 2024 Phoenix Library Management System - Pinnacle Technologies, Inc. All Rights Reserved.