Inductive schemes for the complete classification of affine hypersurfaces with parallel second fundamental form (Q536145)

The topic of this paper is interesting: The author investigates Blaschke hypersurfaces in any dimension \(n \geq 2\) that have a parallel cubic form with respect to the induced connection. Following a result of \textit{N. Bokan, K. Nomizu} and \textit{U. Simon} [``Affine hypersurfaces with parallel cubic forms'', Tôhoku Math. J., II. Ser. 42, No.~1, 101--108 (1990; Zbl 0696.53006)], this class forms a subclass of the large class of improper affine spheres. For \(n = 2\), the solution of the classification is due to \textit{K. Nomizu} and \textit{U. Pinkall} [``Cayley surfaces in affine differential geometry'', Tôhoku Math. J., II. Ser. 41, No.~4, 589--596 (1989; Zbl 0684.53011)]; for \(n = 3\), due to \textit{L. Vrancken} [``Affine higher order parallel hypersurfaces'', Ann. Fac. Sci. Toulouse, V. Sér., Math. 9, No.~3, 341--353 (1988; Zbl 0687.53011)]; for \(n = 4,5\), due to the author [``On affine hypersurfaces with parallel second fundamental form'', Tôhoku Math. J., II. Ser. 54, No.~4, 495--512 (2002; Zbl 1033.53009); ``Classification of five dimensional hypersurfaces with affine normal parallel cubic form'', Beitr. Algebra Geom. 44, No.~2, 511--524 (2003; Zbl 1050.53015)]. Up to \(n = 5\), the known classifications are explicit. The main result of the present paper is that there exist inductive schemes such that the classification in higher dimension can be given by a reductive process to lower dimensions; this classification does not give explicit representations of the hypersurfaces. Due to the author, the main tool is a Monge-Ampère equation that, for a given special coordinate system in affine space, locally characterizes improper affine spheres in terms of this PDE. The author uses this PDE to characterize all improper affine spheres with arbitrary Blaschke metric; the reviewer does know a published proof for this characterization only in the case of locally strongly convex improper affine spheres (see, e.g., the monograph by \textit{A.-M. Li, U. Simon} and \textit{G. Zhao} [Global affine differential geometry of hypersurfaces. Berlin: Walter de Gruyter (1993; Zbl 0808.53002)]). Reviewer's remarks: {\parindent=6mm \begin{itemize}\item[(1.)] Recently, \textit{Z. Hu, C. Li, H. Li} and \textit{L. Vrancken} [``Lorentzian affine hypersurfaces with parallel cubic form'', Result. Math. 59, No.~3--4, 577--620 (2011; Zbl 1223.53010)] gave an explicit classification of all locally strongly convex Blaschke hypersurfaces that have a parallel cubic form with respect to the Levi-Civita connection of the Blaschke metric, \(n \geq 2\). This class is a subclass of the class of hyperbolic affine spheres; they also gave a similar classification in case the Blaschke metric is Lorentzian. For arbitrary Blaschke metrics, the classification is in progress. To reduce the higher-dimensional classification to lower dimensions, the authors use Calabi's decomposition of hyperbolic affine spheres. \item[(2.)] In his paper, Gigena uses a terminology that is different from the standard terminology in the literature, e.g., he calls the cubic form the second fundamental form. In his paper, he does not say which connection he is using (there are three in each relative geometry; see, e.g., the monograph by Li, Simon and Zhao [loc. cit.]), the reader has to find it out by himself. \item[(3.)] In August 2011, the reviewer sent an e-mail to the author asking where in the literature a proof is given of the characterization of improper affine spheres with indefinite metric in terms of a Monge-Ampère equation; until December 2011 there has been no reply to this request. \end{itemize}}