refactor(ml/incr): convert loss function names to kebab-case

[Ashutosh-1304]

Resolves #13359

Description

This pull request refactors loss function option names in @stdlib/ml/incr/binary-classification and @stdlib/ml/incr/sgd-regression from camelCase to kebab-case to conform with stdlib naming conventions:

• @stdlib/ml/incr/binary-classification:
  • modifiedHuber $\rightarrow$ modified-huber
  • squaredHinge $\rightarrow$ squared-hinge
• @stdlib/ml/incr/sgd-regression:
  • epsilonInsensitive $\rightarrow$ epsilon-insensitive
  • squaredError $\rightarrow$ squared-error

This includes updates to core module options, default configurations, TypeScript definitions, validation checks, tests, REPL files, and package documentation.

Related Issues

This pull request has the following related issues:

• [BUG]: usage of camelcase for loss functions in ml/incr/* #13359

Questions

No.

Other

No.

Checklist

• Read, understood, and followed the [contributing guidelines][contributing].

AI Assistance

• Yes
• No

@stdlib-js/reviewers