$($d$)(4$ points$)$ What are some limitations of using accuracy as an evaluation metric for

classification models? Explain at least two different limitations