Schur Positivity and Optimal Cophase Root Configuration

Hello! Welcome back.

In our previous lesson, we identified the paper's candidate for the optimal configuration of characteristic roots: the "cophase" configuration, where all roots coincide on the boundary of the allowed disk, represented by a characteristic polynomial like $(z-r{)}^n$.

Today, we will connect the dots and see precisely how the algebraic tools we've studied—specifically Schur polynomials—are used to prove that this cophase configuration is indeed optimal. This lesson addresses the core of the paper's argument.

Lesson Goal: By the end of this lesson, you will be able to connect the Schur positivity argument to the conclusion that the cophase root configuration is optimal for the interpolation problem.

1. Recapping the Strategy: From Amplitudes to Polynomials

Let's quickly recall the overall proof strategy established in earlier modules:

1. The Goal: Maximize the amplitude $|x_t|$ of the solution to a difference equation.
2. The Constraint: The characteristic roots $z_1,\dots ,z_n$ must lie in the disk $|z_k|\le r$.
3. The Reformulation: The problem was reframed as maximizing a quantity $M_t$, which is a weighted sum of the absolute values of the coefficients of an interpolating polynomial, ${\psi }_t(z)$. The maximization is over the choice of roots $z_1,\dots ,z_n$.

The crucial step, and the paper's key insight, is to find an explicit expression for the coefficients of this interpolating polynomial in terms of the roots $z_1,\dots ,z_n$.

2. The Key Insight: Coefficients as Schur Polynomials

The authors of the paper derive a remarkable identity. They show that the coefficients of the interpolating polynomial ${\psi }_t(z)$ are, up to a sign, specific Schur polynomials evaluated at the characteristic roots.

Let's see this directly in the paper.

As you've just read, the proposition states:

where $s_{(a|b)}$ denotes the Schur polynomial for a hook partition.

This is a pivotal result. The problem of maximizing $M_t={\sum }_{k=1}^n{w}_k\left|\frac{{\psi }_t^{(k-1)}(0)}{(k-1)!}\right|$ over all roots $z_{1:n}\in D(r{)}^n$ now becomes:

(Note the index shift from $k$ to $k-1$ for the derivative order).

The problem has been transformed into maximizing a sum of absolute values of Schur polynomials. This is where "Schur positivity" comes into play.

3. The Role of Schur Positivity

What exactly is Schur positivity, and why is it useful here?

A symmetric function is called Schur-positive (or s-positive) if, when written as a linear combination of Schur polynomials, all the coefficients are non-negative. While interesting, this is not the property we need directly.

The crucial property for our maximization problem is m-positivity. A symmetric function $g$ is m-positive if its expansion in the basis of monomial symmetric functions ($m_{\lambda }$) has non-negative coefficients.

The connection is that Schur functions are themselves m-positive.

From the reading, we have two key takeaways:

1. The definition $s_{\lambda }={\sum }_{\mu }{K}_{\lambda \mu }{m}_{\mu }$ shows that Schur functions are m-positive because the Kostka numbers $K_{\lambda \mu }$ are non-negative integers (they count semi-standard Young tableaux).
2. Fact 3.1 states that if a homogeneous symmetric function $g$ is m-positive, the maximum of $|g(z_1,\dots ,z_n)|$ over the polydisc $D_n(r_1,\dots ,r_n)$ is attained when the roots are on the boundary with the same phase: $z_k=r_k{e}^{i\theta }$. For our simpler case of a single disk $D(r{)}^n$, this means $z_k=r{e}^{i\theta }$ for all $k$.

This is exactly the property we need to solve the maximization problem.

4. The Final Argument

We can now assemble the complete proof. The paper does this in the lines immediately following Proposition 4.2.

1. Start with the goal: We want to find ${\max }_{z_{1:n}\in D(r{)}^n}{M}_t$.
2. Substitute the coefficients: Using Proposition 4.2, we rewrite $M_t$ as the sum of absolute values of Schur polynomials:
   $M_t={\sup }_{z_{1:n}\in D(r{)}^n}{\sum }_{k=1}^n{w}_k|s_{(t-n|n-k)}(z_1,\dots ,z_n)|$
3. Apply m-positivity: Since each Schur polynomial $s_{\lambda }$ is m-positive, we know from Fact 3.1 that the maximum value of each term $|s_{\lambda }(z_{1:n})|$ is achieved when all roots are on the boundary circle with a common phase, i.e., $z_k=r{e}^{i\theta }$.
4. Distribute the supremum: Because all terms in the sum are non-negative and are maximized by the same configuration of roots, the supremum of the sum is the sum of the suprema.
   $\max M_t={\sum }_{k=1}^n{w}_k{\sup }_{z_{1:n}\in D(r{)}^n}|s_{(t-n|n-k)}(z_1,\dots ,z_n)|={\sum }_{k=1}^n{w}_k{s}_{(t-n|n-k)}(r,\dots ,r)$
5. Conclusion: The maximum amplitude is achieved for the cophase root configuration $z_1=z_2=\cdots =z_n=r{e}^{i\theta }$.

The "Schur positivity structure" mentioned in the abstract is precisely this realization: the problem's core quantities are Schur polynomials, whose inherent m-positivity property makes the otherwise difficult maximization problem tractable.

5. Deeper Dive: Why are Schur Polynomials "Positive"?

The properties of Schur polynomials might seem like a convenient mathematical trick. However, their "positivity" and related concavity properties are deep results with connections to representation theory, algebraic geometry, and combinatorics. Your background suggests you might appreciate a glimpse into this deeper theory.

The lecture by Fields Medalist June Huh provides a modern perspective on why Schur polynomials exhibit such strong structural properties.

As the video explains, the non-negativity of the coefficients in the monomial expansion of a Schur polynomial is not just a combinatorial accident; it's a consequence of them being dimensions of vector spaces (weight spaces in a representation). The log-concavity result presented by Huh is a more recent and much stronger statement, revealing a profound geometric structure underlying these polynomials. The paper's argument relies on the simpler m-positivity, but this deeper structure is the ultimate reason why the argument works.

Conclusion

In this lesson, we have traced the main argument of the paper, connecting the amplitude maximization problem to the theory of symmetric functions.

Key Takeaways:

• The coefficients of the key interpolating polynomial in the problem are, up to a sign, Schur polynomials of the characteristic roots.
• Schur polynomials are m-positive, meaning their expansion in the monomial symmetric basis has non-negative coefficients (Kostka numbers).
• A crucial property of m-positive homogeneous symmetric functions is that their modulus is maximized over a disk when all variables are on the boundary with a common phase.
• This property allows for a direct solution to the maximization problem, proving that the cophase root configuration is optimal.
• The phrase "Schur positivity structure" refers to the identification of Schur polynomials as the core components of the problem, which unlocks their powerful properties for the proof.

Next Lesson Preview:

We have now established why the cophase configuration is optimal. In the next lesson, we will put everything together to state the paper's main result formally and interpret its consequences, particularly for the peak amplitude over all time, ${\sup }_{t\ge n}|x_t|$.