Making Convex.jl DPP compliant

Hi Team

I read about [the project](https://github.com/jump-dev/GSOC2020/blob/master/ideas-list.md#optimization-problem-differentiation) to enable `JuMP.jl` differentiate solution of problem w.r.t. its parameters.

CVXPY implemented this ability and [the article](http://web.stanford.edu/~boyd/papers/pdf/diff_cvxpy.pdf) accompanying this feature describes a new [grammar named DPP](https://www.cvxpy.org/tutorial/advanced/index.html#disciplined-parametrized-programming), subset of the DCP system.

Does `Convex.jl` defines a `parameter` class like a variable or constraint? (I didn't find any implementation [like this one](https://github.com/cvxgrp/cvxpy/blob/master/cvxpy/expressions/constants/parameter.py))

Start by reading the linked DPP documentation and CVXPY's parameter implementation, then inspect Convex.jl's handling of variables and constraints to determine whether parameters already have an equivalent. Define the required parameter model and DPP compliance criteria before identifying the affected implementation and tests; the issue does not name specific files or tests.

Written by the indexing model from the issue text.