A Novel Method for Analysing Racial Bias: Appendix: Hyperparameter Sensitivity

14 May 2024

This paper is available on arxiv under CC 4.0 license.

Authors:

(1) Muhammed Yusuf Kocyigit, Boston University;

(2) Anietie Andy, University of Pennsylvania;

(3) Derry Wijaya, Boston University.

Table of Links

Appendix: Hyperparameter Sensitivity

We conduct our experiments with an array of hyperparameters to observe the effects of the sampling size and the negative sampling ration and observe that the results are relatively stable. Below are the results for k=[500K, 1M] and n=[1,4,10,20]

Table 2: Caption

Table 3: Kolmogorov–Smirnov two-sample test for the African American Embeddings in Table 5

Table 4: Kolmogorov–Smirnov two-sample test for the White American Embeddings in Table 5

Figure 9: Correlation heatmap for k=500K, n=1. Left-White Americans, Right-African American

Figure 10: Correlation heatmap for k=500K, n=4. Left-White Americans, Right-African American

Figure 12: Correlation heatmap for k=500K, n=20. Left-White Americans, Right-African American

Figure 13: Correlation heatmap for k=1M, n=1. Left-White Americans, Right-African American

Figure 15: Correlation heatmap for k=1M, n=10. Left-White Americans, Right-African American

Figure 16: Correlation heatmap for k=1M, n=120. Left-White Americans, Right-African American