The Setting
For an optimal Gaussian partition, the active cells have nonzero centroids. Let be the active cells of an optimal Gaussian partition in Euclidean space , with its usual inner product; write for standard Gaussian measure and use the notation in (1): (1)
Scalar Data
A scalar probability datum of size is a pair of real sequences satisfying (2), where is the coefficient specified there: (2) For , put and . The scalar feasible set is the set of scalar probability data satisfying the two determinant constraints (3): (3)
For a real parameter , the scalar charge is the quadratic polynomial in (4), and its charge threshold is the scalar given there: (4) Budgeted ordered radii with parameter are real sequences whose radii satisfy the ordering, positivity, cap and squared normalization in (2), and whose excesses satisfy (5): (5)✦ For real parameters and , the clipping coefficient and smaller budget root are real numbers, and the terminal charge budget is a real function, defined in (6). (6) An admissible scalar regime is a pair consisting of a real parameter and an integer satisfying (7); write within that display: (7)
Normalized Configurations
The unnormalized Gaussian density , Gaussian normalizing constant and unnormalized upper tail are the functions and constant in (8): (8) Let and denote the standard Gaussian density and distribution function. The Gaussian upper tail , inverse Mills ratio , inverse Mills derivative in the paper’s argument, and Gaussian loss are the real functions defined by (9); the reflected argument is : (9)
For a nonzero vector and , the negative projection coefficient is the real number and the orthogonal residual is the vector in defined by (10). For normalized vectors, use the abbreviations in the same display: (10) For centroids normalized as in (1), is the unnormalized residual divided by .
A normalized obtuse configuration with labels in is a family of vectors and real cell probabilities whose lengths form a scalar probability datum, and which satisfy (11) and (12): (11) (12) When , a normalized obtuse configuration also satisfies the four-cell cap and loss conditions (13), for every : (13)
The Claim
Theorem 1. Let . For budgeted ordered radii with parameter , the first constraint in (3) implies (14)✦ For budgeted ordered radii in an admissible scalar regime , suppose that and . The tail cap, minimum bound and third-radius estimate are, respectively, (15)✦ (16)✦ (17)✦ For real numbers with and , one has (18)✦ The scalar feasible set satisfies (19)✦ Every normalized obtuse configuration in has (20)✦
In the paper, the result reads: for , positive ordered radii with and , and probabilities summing to one, cannot satisfy the three bounds of Lemma 3.1 and the two determinant constraints of Corollary 4.2, with a scalar threshold and the estimates , for , and preceding the contradiction to the second determinant constraint.
What Is Checked
The scalar assertion (19) is stated over the finite real sequences in (2) and (3); the radius estimates (14), (15), (16) and (17) concern budgeted ordered radii and admissible scalar regimes. The determinant assertion (18) concerns two real scalars, and the cell-count assertion (20) concerns normalized obtuse configurations with the residual-pair and four-cell conditions above.