Rice researchers found two machine learning models widely used for immunotherapy research did not correct for bias present in the publicly available data used to train the models, which appears to favor higher-income populations.
Widely used machine learning models reproduce dataset bias in Rice study
High-income communities overrepresented in relevant datasets for immunotherapy research
Link: https://csweb.rice.edu/news/rice-researchers-find-bias-healthcare-data-hope-spur-greater-inclusion
Body