Convex Repeated Games and Fenchel Duality

نویسندگان

  • Shai Shalev-Shwartz
  • Yoram Singer
چکیده

We describe and analyze an algorithmic framework for playing convex repeatedgames. In each trial of the repeated game, the first player predicts a vector andthen the second player responds with a loss function over the vector. Based on ageneralization of Fenchel duality, we derive an algorithmic framework for the firstplayer and analyze the player’s regret. We then use our algorithmic framework andits corresponding regret analysis for online learning problems and for boosting.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Stable and Total Fenchel Duality for Convex Optimization Problems in Locally Convex Spaces

STABLE AND TOTAL FENCHEL DUALITY FOR CONVEX OPTIMIZATION PROBLEMS IN LOCALLY CONVEX SPACES∗ CHONG LI† , DONGHUI FANG‡ , GENARO LÓPEZ§ , AND MARCO A. LÓPEZ¶ Abstract. We consider the optimization problem (PA) infx∈X{f(x) + g(Ax)} where f and g are proper convex functions defined on locally convex Hausdorff topological vector spaces X and Y , respectively, and A is a linear operator from X to Y ....

متن کامل

Asymptotic closure condition and Fenchel duality for DC optimization problems in locally convex spaces

We consider the DC optimization problem (P) inf x∈X {(f1(x) − f2(x)) + (g1(Ax) − g2(Ax))}, where f1, f2, g1 and g2 are proper convex functions defined on locally convex Hausdorff topological vector spaces X and Y respectively, and A is a linear continuous operator from X to Y . Adopting the standard convexification technique, a Fenchel dual problem of (P) is given. By using properties of the ep...

متن کامل

Stable and Total Fenchel Duality for DC Optimization Problems in Locally Convex Spaces

We consider the DC (difference of two convex functions) optimization problem ðPÞ inf x∈Xfðf 1ðxÞ− f 2ðxÞÞ þ ðg1ðAxÞ− g2ðAxÞÞg, where f 1, f 2, g1, and g2 are proper convex functions defined on locally convex Hausdorff topological vector spaces X and Y , and A is a linear continuous operator from X to Y . Adopting different tactics, two types of the Fenchel dual problems of ðPÞ are given. By usi...

متن کامل

Farkas-type results and duality for DC programs with convex constraints

In this paper, we are interested in new versions of Farkas lemmas for systems involving convex and DC-inequalities. These versions extend well-known Farkas-type results published recently, which were used as main tools in the study of convex optimization problems. The results are used to derive several strong duality results such as: Lagrange, Fenchel-Lagrange or Toland-Fenchel-Lagrange duality...

متن کامل

Symbolic Computation of Fenchel Conjugates

Convex optimization is a branch of mathematics dealing with nonlinear optimization problems with additional geometric structure. This area has been the focus of considerable recent research due to the fact that convex optimization problems are scalable and can be efficiently solved by interior-point methods. Over the last ten years or so, convex optimization has found new applications in many a...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2006