Disclosure Determinants of Blockchain Crowdfunding Performance for Sustainable Smart City Financing: An Explainable Optuna-Optimized Machine Learning Approach

Smart cities present a transformative paradigm for urban development, yet securing sustainable financing remains a critical challenge. While traditional funding mechanisms struggle with scalability limitations, FinTech innovations like Initial Coin Offerings (ICOs) have emerged as a viable alternative. Leveraging blockchain technology, ICOs enable decentralized capital raising through token sales, offering transparency and global investor access. However, their effectiveness is compromised by market volatility, information asymmetry, and the absence of reliable predictive frameworks. This study addresses these limitations by developing an explainable hybrid machine learning model that combines: (1) Light Gradient Boosting Machine (LGBM) for efficient feature selection through histogram-based learning, (2) Optuna-optimized Extremely Randomized Trees regression that mitigates overfitting via enhanced randomization while excelling with noisy financial data, and (3) interpretability tools including SHAP values and feature importance analysis. Optuna's automated hyperparameter optimization further enhances computational efficiency, enabling robust predictions of post-ICO returns. The proposed model demonstrates superior predictive performance (R²=0.814, MSE=0.005, MAE=0.051), significantly outperforming both linear regression and state-of-the-art ML models. Key findings identify token supply (63% predictive power) as negatively correlated with returns - reflecting dilution effects and investor perceptions of scarcity- while fundraising success (15%) and Bitcoin returns (8%) show positive influences. These results provide practical guidance for investors and regulators, while establishing ICOs as a potential sustainable financing mechanism for smart city initiatives. The study contributes both methodologically through its optimized hybrid architecture and practically by enhancing decision-making in blockchain-based urban development financing.

Authors

Institutions

Publication Details

Journal
Journal of Posthumanism
Published
2025-07-04
DOI
https://doi.org/10.63332/joph.v5i7.2843
Citations
1
Primary Topic
FinTech, Crowdfunding, Digital Finance
Type
article
Field-Weighted Citation Impact
2.28

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Disclosure Determinants of Blockchain Crowdfunding Performance for Sustainable Smart City Financing: An Explainable Optuna-Optimized Machine Learning Approach

Gasser G. Ali, Zakaria Yahia
1 citations
Journal of Posthumanism
FinTech, Crowdfunding, Digital Finance
2.28
article

Disclosure Determinants of Blockchain Crowdfunding Performance for Sustainable Smart City Financing: An Explainable Optuna-Optimized Machine Learning Approach

Gasser G. Ali, Zakaria Yahia
article en
1 citations

Abstract

Smart cities present a transformative paradigm for urban development, yet securing sustainable financing remains a critical challenge. While traditional funding mechanisms struggle with scalability limitations, FinTech innovations like Initial Coin Offerings (ICOs) have emerged as a viable alternative. Leveraging blockchain technology, ICOs enable decentralized capital raising through token sales, offering transparency and global investor access. However, their effectiveness is compromised by market volatility, information asymmetry, and the absence of reliable predictive frameworks. This study addresses these limitations by developing an explainable hybrid machine learning model that combines: (1) Light Gradient Boosting Machine (LGBM) for efficient feature selection through histogram-based learning, (2) Optuna-optimized Extremely Randomized Trees regression that mitigates overfitting via enhanced randomization while excelling with noisy financial data, and (3) interpretability tools including SHAP values and feature importance analysis. Optuna's automated hyperparameter optimization further enhances computational efficiency, enabling robust predictions of post-ICO returns. The proposed model demonstrates superior predictive performance (R²=0.814, MSE=0.005, MAE=0.051), significantly outperforming both linear regression and state-of-the-art ML models. Key findings identify token supply (63% predictive power) as negatively correlated with returns - reflecting dilution effects and investor perceptions of scarcity- while fundraising success (15%) and Bitcoin returns (8%) show positive influences. These results provide practical guidance for investors and regulators, while establishing ICOs as a potential sustainable financing mechanism for smart city initiatives. The study contributes both methodologically through its optimized hybrid architecture and practically by enhancing decision-making in blockchain-based urban development financing.

Journal of PosthumanismVol. 5(7)
Prince Sattam Bin Abdulaziz University (SA), Egypt-Japan University of Science and Technology (EG), University of Business and Technology (SA)
Prince Sattam bin Abdulaziz University
Openalex Percentile: Top 11%
FinTech, Crowdfunding, Digital Finance
2.28
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.