The chapters are split into sections, which, in turn, are split into subsections enumerated by two numbers: the first stands for the number of the section while the second for the number ofthe subsection itself. The same numeration is used for all kinds of statements and formulas. If we refer to statements or formulas in other chapters, we use triple numeration where the first number stands for the chapter and the other two have the same sense. The results presented in this book were discussed on the seminars at the Institute of Mathematics of Ukrainian Academy ofSciences, at the Steklov Mathematical Institute of the Academy of Sciences of the USSR, at Moscow and Tbilisi State Universities. I am deeply grateful to the heads of these seminars Professors V. K. Dzyadyk, N. P. Kor- neichuk, S. B. Stechkin, P. L. U1yanov, and L. V. Zhizhiashvili as well as to the mem- bers ofthese seminars that took an active part in the discussions. In TRODUCTIon It is well known for many years that every 21t -periodic summable function f(x) can be associated in a one-to-one manner with its Fourier series (1. 1) Slfl where I It = - f f(t)cosktdt 1t -It and I It - f f(t)sinktdt. 1t -It Therefore, if for approximation of a given function f(*), it is necessary to construct a sequence ofpolynomials Pn (.

The chapters are split into sections, which, in turn, are split into subsections enumerated by two numbers: the first stands for the number of the section while the second for the number ofthe subsection itself. The same numeration is used for all kinds of statements and formulas. If we refer to statements or formulas in other chapters, we use triple numeration where the first number stands for the chapter and the other two have the same sense. The results presented in this book were discussed on the seminars at the Institute of Mathematics of Ukrainian Academy ofSciences, at the Steklov Mathematical Institute of the Academy of Sciences of the USSR, at Moscow and Tbilisi State Universities. I am deeply grateful to the heads of these seminars Professors V. K. Dzyadyk, N. P. Kor- neichuk, S. B. Stechkin, P. L. U1yanov, and L. V. Zhizhiashvili as well as to the mem- bers ofthese seminars that took an active part in the discussions. In TRODUCTIon It is well known for many years that every 21t -periodic summable function f(x) can be associated in a one-to-one manner with its Fourier series (1. 1) Slfl where I It = - f f(t)cosktdt 1t -It and I It - f f(t)sinktdt. 1t -It Therefore, if for approximation of a given function f(*), it is necessary to construct a sequence ofpolynomials Pn (.

The theory of approximation of functions is one of the central branches in mathematical analysis and has been developed over a number of decades. This monograph deals with a series of problems related to one of the directions of the theory, namely, the approximation of periodic functions by trigonometric polynomials generated by linear methods of summation of Fourier series. More specific, the following linear methods are investigated: classical methods of Fourier, Fejir, Riesz, and Roginski. For these methods the so-called Kolmogorov-Nikol'skii problem is considered, which consists of finding exact and asymptotically exact qualities for the upper bounds of deviations of polynomials generated by given linear methods on given classes of 2?-periodic functions. Much attention is also given to the multidimensional case.
The material presented in this monograph did not lose its importance since the publication of the Russian edition (1981). Moreover, new material has been added and several corrections were made. In this field of mathematics numerous deep results were obtained, many important and complicated problems were solved, and new methods were developed, which can be extremely useful for many mathematicians. All principle problems considered in this monograph are given in the final form, i.e. in the form of exact asymptotic equalities, and, therefore, retain their importance and interest for a long time.