Claim Missing Document
Check
Articles

Found 1 Documents
Search
Journal : JITU : Journal Informatic Technology And Communication

Implementasi Hierarchical Clustering untuk Analisis FDMC Narrative Crypto Berbasis Web M. Fathir Adha; Hendrik Fery Herdiatmoko
JITU Vol 10 No 1 (2026)
Publisher : Universitas Boyolali

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.36596/jitu.v10i1.2255

Abstract

This study implements the Hierarchical Clustering algorithm with Ward linkage and Euclidean distance methods to analyze 26 crypto narratives based on the Fully Diluted Market Cap (FDMC) metric. Using a hybrid method that integrates Waterfall, Cross-Industry Standard Process for Data Mining (CRISP-DM), and Knowledge Discovery in Databases (KDD), data was obtained from the CoinGecko API, manually clustered, and aggregated per narrative. Pre-processing involved logarithmic transformation (log-10) and Z-Score normalization to address power-law distributions and outliers, resulting in a more stable cluster structure. The clustering results mapped the market into five clusters: Bluechip (L1 with FDMC $2.76T), Growth (PAY, MEME, CEX, DEX, DeFi totaling $468.22B), Growth (AI, DePIN, DAO, L2, RWA, ORC, GameFi, XCH, DID, PRC, LST with $192.91B), Speculative (NFT, MET, SocialFi, BTC Eco, W3I with $17.55B), and Speculative (LPD, GambleFi, FTO, SEC with $2.34B). The model was validated with a Silhouette Score of 0.650 and a Cophenetic Correlation Coefficient of 0.647, indicating cohesive and representative clusters. A web-based implementation using Django, D3.js, and Chart.js provides interactive visualizations and portfolio recommendations. Contributions include a novel fundamental valuation approach, an adaptive clustering model, and practical analytical tools for investors, with potential expansion to multidimensional metrics in the future.