No abstract uploaded!

The stochastic gradient (SG) method can quickly solve a problem with a large number of components in the objective, or a stochastic optimization problem, to a moderate accuracy. The block coordinate descent/update (BCD) method, on the other hand, can quickly solve problems with multiple (blocks of) variables. This paper introduces a method that combines the great features of SG and BCD for problems with many components in the objective and with multiple (blocks of) variables. This paper proposes a block SG (BSG) method for both convex and nonconvex programs. BSG generalizes SG by updating all the blocks of variables in the Gauss–Seidel type (updating the current block depends on the previously updated block), in either a fixed or randomly shuffled order. Although BSG has slightly more work at each iteration, it typically outperforms SG because of BSG’s Gauss–Seidel updates and larger step sizes, the latter of which are determined by the smaller per-block Lipschitz constants. The convergence of BSG is established for both convex and nonconvex cases. In the convex case, BSG has the same order of convergence rate as SG. In the nonconvex case, its convergence is established in terms of the expected violation of a first-order optimality condition. In both cases our analysis is nontrivial since the typical unbiasedness assumption no longer holds. BSG is numerically evaluated on the following problems: stochastic least squares and logistic regression, which are convex, and low-rank tensor recovery and bilinear logistic regression, which are nonconvex. On the convex problems, BSG performed significantly better than SG. On the nonconvex problems, BSG significantly outperformed the deterministic BCD method because the latter tends to stagnate early near local minimizers. Overall, BSG inherits the benefits of both SG approximation and block coordinate updates and is especially useful for solving large-scale nonconvex problems.

We perform a numerical study of the heat transfer and flow structure of Rayleigh-Benard (RB) convection in (in most cases regular) porous media, which are comprised of circular, solid obstacles located on a square lattice. This study is focused on the role of porosity in the flow properties during the transition process from the traditional RB convection with (so no obstacles included) to Darcy-type porous-media convection with approaching 0. Simulations are carried out in a cell with unity aspect ratio, for Rayleigh number from to and varying porosities , at a fixed Prandtl number , and we restrict ourselves to the two-dimensional case. For fixed , the Nusselt number is found to vary non-monotonically as a function of ; namely, with decreasing , it first increases, before it decreases for approaching 0. The non-monotonic behaviour of originates from two competing effects of the porous structure on the heat transfer. On the one hand, the flow coherence is enhanced in the porous media, which is beneficial for the heat transfer. On the other hand, the convection is slowed down by the enhanced resistance due to the porous structure, leading to heat transfer reduction. For fixed , depending on , two different heat transfer regimes are identified, with different effective power-law behaviours of versus , namely a steep one for low when viscosity dominates, and the standard classical one for large . The scaling crossover occurs when the thermal boundary layer thickness and the pore scale are comparable. The influences of the porous structure on the temperature and velocity fluctuations, convective heat flux and energy dissipation rates are analysed, further demonstrating the competing effects of the porous structure to enhance or reduce the heat transfer.

No abstract uploaded!

For an odd prime p, we prove the p-part of the Birch and Swinnerton-Dyer formula for semistable elliptic curves of analytic rank one.