Academic Journal
A comprehensive study of algorithmic bias : fairness in AI, ethical and social implications in real world applications
| Τίτλος: | A comprehensive study of algorithmic bias : fairness in AI, ethical and social implications in real world applications |
|---|---|
| Συγγραφείς: | Jain, Lakshitha Rikhab Chand |
| Πηγή: | Theses |
| Στοιχεία εκδότη: | LOUIS |
| Έτος έκδοσης: | 2023 |
| Θεματικοί όροι: | Computer algorithms--Design, Artificial intelligence--Social aspects, Discrimination--Prevention |
| Περιγραφή: | Algorithmic bias, also known as algorithmic discrimination or AI bias, is a pervasive phenomenon marked by the systematic and unjust treatment of specific demographic groups based on attributes like race, gender, age or other distinguishing characteristics. The widespread adoption of artificial intelligence (AI) and machine learning algorithms across various domains has amplified concerns surrounding algorithmic bias, which has the potential to perpetuate discrimination and deepen societal inequalities. This thesis embarks on a comprehensive exploration of algorithmic bias, delving into its multifaceted nature. It dissects various forms of bias, including selection bias, confirmation bias, and measurement bias, shedding light on the catalysts driving these biases. Factors such as data quality, algorithm design choices, and institutional biases are scrutinized to unveil the roots of algorithmic bias. Importantly, this study underscores the detrimental repercussions of algorithmic bias, which span from reinforcing social disparities to hindering societal progress. With a steadfast commitment to impartiality and equity, the research aims to unearth the causes and effects of algorithmic biases. Furthermore, it endeavors to propose effective interventions that foster social advancement and extend benefits to individuals from diverse backgrounds. By doing so, this thesis contributes to the ongoing evolution of AI systems towards a more equitable and just future. |
| Τύπος εγγράφου: | text |
| Περιγραφή αρχείου: | application/pdf |
| Γλώσσα: | unknown |
| Διαθεσιμότητα: | https://louis.uah.edu/uah-theses/624 https://louis.uah.edu/context/uah-theses/article/1624/viewcontent/jain_10898.pdf |
| Αριθμός Καταχώρησης: | edsbas.1D8AFEF3 |
| Βάση Δεδομένων: | BASE |
καταχωρήστε σχόλιο πρώτοι!