Register Login

Journal of Computational Social Dynamics

Search
PDF

Published

2022-10-03

Issue

Vol. 7 No. 4 (2022): Journal of Computational Social Dynamics-4th-Q

Section

Articles

How to Cite

Sharma, A., & Joshi, T. (2022). Big Data for Central Banks: Exploring Applications in Monetary Policy, Financial Stability Monitoring and Bank Supervision. Journal of Computational Social Dynamics, 7(4), 1–10. Retrieved from https://vectoral.org/index.php/JCSD/article/view/41

Big Data for Central Banks: Exploring Applications in Monetary Policy, Financial Stability Monitoring and Bank Supervision

Arjun Sharma

Department of Economics and Finance, Indian Institute of Management Bangalore (IIMB), India 

Tripuresh Joshi

Tula‘s Institute Dehradun

Keywords: Artificial Intelligence, Big Data Analytics, Algorithmic Auditing, Biases in AI Models, Model Fairness, Transparency of Proprietary Models

Abstract

With the rapid adoption of artificial intelligence (AI) and big data analytics across high-impact domains like finance, healthcare and public policy, there are rising concerns about inadvertent biases and discrimination arising from opaque algorithms. Several instances have already emerged of AI systems unfairly disadvantaging minorities, women and vulnerable social groups due to historical biases in data, poor choice of proxy variables, narrow optimization objectives and lack of diversity. Addressing this critical issue, the field of algorithmic auditing has emerged to provide testing methodologies that can uncover harmful biases in AI models by auditing the entire pipeline from data collection to model development and deployment. Algorithmic auditing combines techniques like causal inference, adversarial testing, counterfactual explanation, and meta-learning to assess model fairness, explain model predictions, detect bias in training data and ultimately redesign systems to mitigate sources of unfairness. However, realizing the potential of algorithmic auditing requires overcoming key challenges around limited transparency of proprietary models, unrepresentative datasets, interpretability versus accuracy tradeoffs and the need for multidisciplinary teams combining technical and social science expertise. Further work is also needed on organizational aspects like developing auditing frameworks tailored to different application domains, integrating audits within model development workflows, and coordinating audits across stakeholders involved in AI systems. Overall, algorithmic auditing is a promising paradigm but still in nascent stages. Advancing frameworks, best practices and inclusive governance models for auditing will be crucial to steer AI systems towards fairness and address the growing risk of algorithmic biases.

Author Biography

Tripuresh Joshi, Tula‘s Institute Dehradun

 

 

 

Current Issue

Information

Submission Guidelines for the Journal of Computational Social Dynamics General Guidelines:

  1. Article Types: We accept a variety of article types including Original Research Articles, Review Papers, Short Reports, and Opinion Pieces.

  2. Language: Manuscripts must be written in English, and should be clear, concise, and free of jargon.

  3. Length: There is no strict word limit, but Original Research Articles should generally be between 4,000 and 10,000 words, including references.

  4. Formatting: Manuscripts should be submitted as a single Word or LaTeX document, formatted in double-spaced 12-point font.

  5. Abstract: An abstract of 250-300 words is required for all submissions, summarizing the key objectives, methods, findings, and implications of the research.

Manuscript Structure:

  • Title Page: Must include the title, author names, affiliations, and contact information for the corresponding author.

  • Introduction: Provide background, state the problem, and outline objectives.

  • Methods: Detail the methods and tools used in your research.

  • Results: Present the findings of your study.

  • Discussion: Interpret results, discuss their impact, and suggest future research directions.

  • Conclusion: Summarize the main points of the paper.

  • References: Citations should be formatted according to the APA/IEEE styles.

Ethical Guidelines:

Authors must ensure that their research adheres to ethical standards and that informed consent has been obtained for any human or animal subjects involved.

Plagiarism:

The Journal of Computational Social Dynamics uses plagiarism detection software. Any manuscript found to include plagiarized content will be rejected.

Peer Review Process:

All submissions undergo a double-blind peer review process. Authors can expect initial feedback within six to eight weeks.

Submission Process:
  1. Prepare your manuscript according to the guidelines.
  2. Submit your manuscript through our online portal at.
  3. After submission, you will receive a confirmation email with a manuscript ID for tracking purposes.
Revisions:

If revisions are requested, authors must resubmit the revised manuscript within the specified timeframe.

Publication Fees:

The journal has a nominal publication fee, waived for authors who lack funding. Details are available on our website.

For any questions or concerns, please contact the editorial office at editorial@vectoral.org. Thank you for considering the Journal of Computational Social Dynamics as a venue for your research.

Published by Vectoral Publishing, the Journal of Computational Social Dynamics maintains high standards in peer-review and scientific inquiry. Authors who wish to submit manuscripts should adhere to the guidelines detailed on our website, www.vectoral.org/jcsd. Vectoral Publishing does not bear any responsibility for the accuracy of information or for any outcomes related to the use of the research published. For editorial correspondence and submissions, please reach out to editorial@vectoral.org.