International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 11 Issue: 06 | Jun 2024
p-ISSN: 2395-0072
www.irjet.net
A Novel of Detecting and Addressing Bias in Artificial Intelligence and Machine Learning: A Multi-Industry Perspective Sukanya Konatam1, Venkata Naga Murali Konatam2, Shravya Konatam3 1Senior Manager of Enterprise Data Governance and Data Science, IT, Vialto Partners, Texas, USA 2Data Architect, IT, Capital One, Texas, USA 3Student, Pre-Med, Nova SouthEastern University, Florida, USA
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Artificial intelligence (AI) and machine
various sectors [1], [2], [3]. AI refers to the simulation of human intelligence processes by machines, especially computer systems [1], [3]. These processes include learning, reasoning, problem-solving, perception, and language understanding [1], [3]. On the other hand, ML is a subset of AI that provides systems the ability to learn and improve from experience without being explicitly programmed [1], [2], [3]. AI and ML have permeated various sectors, revolutionizing processes and systems [1], [2], [3].This research paper will focus on four key sectors where AI/ML has a significant impact: the Hiring Process, Healthcare, Financial Systems, and Self-Driving Cars [3][6]. Surprisingly, these sectors are not immune to biases in AI/ML models, which can have profound implications [4], [5]. The exploration of these biases forms the crux of this research. These technologies have the potential to bring about significant benefits, such as increased efficiency, improved decision-making, and enhanced user experiences [1], [2], [3]. However, they also pose new challenges, such as the risk of biases in AI/ML models [4], [5]. This paper will explore these issues in detail, with a focus on understanding the sources of these biases and how they can be mitigated [4], [5].
learning (ML) have gained widespread popularity and transformative potential in various critical sectors, including healthcare, hiring, self-driving cars, and financial systems. Despite their benefits, the adoption of AI and ML systems introduces biases that can profoundly impact decision-making processes and outcomes. These biases stem from various sources, such as biased training data, algorithmic design, and implementation practices, leading to ethical concerns and consequences. This paper provides a comprehensive exploration of the biases inherent in AI and ML systems and discusses their ethical implications across the aforementioned sectors. Through a systematic review of current literature, we identify the primary sources and types of biases, such as racial, gender, and socioeconomic biases, and analyze their effects on decision-making. Additionally, we examine case studies that illustrate the real-world impact of biased AI systems, shedding light on the critical need for fair AI development. To address these challenges, we propose a set of mitigation strategies aimed at reducing bias and enhancing the fairness and accountability of AI systems. These strategies include improving the diversity and representativeness of training data, implementing algorithmic fairness techniques, conducting regular audits and impact assessments, and fostering interdisciplinary collaboration among AI developers, ethicists, and policymakers. Our findings highlight the urgent need for a holistic approach to AI development that prioritizes ethical considerations and social responsibility. By adopting these mitigation strategies, we can work towards creating AI systems that not only deliver technological advancements but also promote equity, transparency, and trust in critical sectors. Key Words: Autonomous Vehicles, Artificial Intelligence in HealthCare, Bias in Machine Learning, AI accountability, AI in job screening, ML in Finance, Bias in Artificial Intelligence
Fig -1: The mathematical representation of the Mean Squared Error (MSE) loss function used in machine learning models. This function measures the average squared difference between the predicted values (mX_i + c) and the actual values (Y_i), highlighting the importance of minimizing prediction errors to improve model accuracy and reduce biases.
1. Introduction to Artificial Intelligence and Machine Learning in Different Sectors
2. Artificial Intelligence and Machine Learning in Healthcare
Artificial Intelligence (AI) and Machine Learning (ML) are transformative technologies that have revolutionized
Artificial Intelligence (AI) and Machine Learning (ML) are transforming the healthcare industry by offering
© 2024, IRJET
|
Impact Factor value: 8.226
|
ISO 9001:2008 Certified Journal
|
Page 831