Direct multiplicative methods for sparse matrices. Quadratic programming
نویسندگان
چکیده
منابع مشابه
Multiplicative Updates for Nonnegative Quadratic Programming
Many problems in neural computation and statistical learning involve optimizations with nonnegativity constraints. In this article, we study convex problems in quadratic programming where the optimization is confined to an axis-aligned region in the nonnegative orthant. For these problems, we derive multiplicative updates that improve the value of the objective function at each iteration and co...
متن کاملDirect Solvers for Sparse Matrices
There is a vast variety of algorithms associated with each step. The review papers by Duff [16] (see also [15, Chapter 6]) and Heath et al. [27] can serve as excellent reference of various algorithms. Usually steps 1 and 2 involve only the graphs of the matrices, and hence only integer operations. Steps 3 and 4 involve floating-point operations. Step 3 is usually the most time-consuming part, w...
متن کاملA Quadratic Programming Method for Ranking Alternatives Based on Multiplicative and Fuzzy Preference Relations
This paper proposes a quadratic programming method (QPM) for ranking alternatives based on multiplicative preference relations (MPRs) and fuzzy preference relations (FPRs). The proposed QPM can be used for deriving a ranking from either a MPR or a FPR, or a group of MPRs, or a group of FPRs, or their mixtures. The proposed approach is tested and examined with two numerical examples, and compara...
متن کاملQuadratic programming for learning sparse codes
Building on a framework by Daugmann [2] and Pece [9], Harpur and Prager [4, 5, 6] constructed a neural network, which they termed the REC model, capable of discovering sparsely distributed representations by using the principle of redundancy reduction. Olshausen and Field [7, 8] employed a similar technique for efficient coding of natural images and showed how the resulting response functions o...
متن کاملDirect linear time solvers for sparse matrices
In the last couple of years it has been realized that Gaussian elimination for sparse matrices arising from certain elliptic PDEs can be done in O(n log(1/ )) flops, where n is the number of unknowns and is a user-specified tolerance [Chandrasekaran and Gu]. The resulting solver will also be backward stable with an error of O( ). The techniques used to achieve this speedup has some commonality ...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Computer Research and Modeling
سال: 2018
ISSN: 2076-7633,2077-6853
DOI: 10.20537/2076-7633-2018-10-4-407-420