Exponential Taylor Series

André Kowacs
2024-08-23
Abstract:This paper derives a way to express differentiable complex-valued functions as the sum of powers of $(1-e^{\lambda x})$, where $\lambda\in\mathbb{R}$, with an explicit formula for the remainder. This formulation is then used to associate an infinite series to $C^\infty$ functions, which is shown to recover the original function under suitable conditions on the remainder. These results are also used to calculate some infinite series involving Stirling Numbers, as well as providing a few examples.
Classical Analysis and ODEs
What problem does this paper attempt to address?
The problem that this paper attempts to solve is how to represent a complex - valued differentiable function as a power series form of \((1 - e^{\lambda x})\) and provide an explicit formula for the remainder term. Specifically, the author derives a method that can represent a \(C^\infty\) function as an infinite series, which can recover the original function under certain conditions. In addition, these results are also used to calculate some infinite series involving Stirling numbers, and several specific examples are provided. ### Main contributions: 1. **New exponential Taylor series expansion**: The paper generalizes the exponential Taylor series expansion proposed by Michael Ruzhansky and Ville Turunen, extending it from the specific \(\lambda = 2\pi i\) to any \(\lambda\in\mathbb{C}\setminus\{0\}\). 2. **Remainder formula**: The author gives an explicit formula for the remainder term, which makes it possible to study the convergence of the infinite series of \(C^\infty\) functions. 3. **Approximation of periodic functions**: By choosing \(\lambda=\frac{2\pi i}{T}\), periodic functions can be approximated as periodic polynomials, and the error term is also periodic. 4. **Specific applications**: The paper shows through specific examples how to use this new exponential Taylor series expansion to calculate some infinite series involving Stirling numbers. ### Mathematical formulas: - **Main theorem**: \[ a(x)=\sum_{j = 0}^{N - 1}\frac{1}{j!}D_{\lambda,(j)}a(x_0)(e^{\lambda(x - x_0)} - 1)^j+R_N(x,x_0), \] where the remainder term \(R_N(x,x_0)\) is: \[ R_N(x,x_0)=\frac{\lambda}{(N - 1)!}\int_{0}^{1}D_{\lambda,(N)}a((1-\theta)x_0+\theta x)(e^{\lambda(1-\theta)(x - x_0)} - 1)^{N - 1}(x - x_0)d\theta. \] - **Estimation of the remainder term**: \[ |R_N(x,x_0)|\leq\frac{|\lambda|^N}{N!}\sup_{\xi\in[[x_0,x]]}|D_{\lambda,(N)}a(\xi)(e^{\lambda(x-\xi)} - 1)^{N - 1}||x - x_0|. \] - **Multivariable case**: \[ a(x)=\sum_{|\gamma|<N}\frac{1}{\gamma!}D_{\lambda,(\gamma)}a(\tilde{x})(e^{\lambda(x - \tilde{x})} - 1)^\gamma+R_N(x,\tilde{x}), \] where the remainder term \(R_N(x,\tilde{x})\) satisfies: \[ |R_N(x,\tilde{x})|\leq|\lambda|\left[\sum_{|\gamma| = N}\frac{N}{\gamma!}\sup_{y\in Q(x,\tilde{x})}|D_{\lambda,(\gamma)}a(y)|\right]\epsilon(\lambda,\|x - \tilde{x}\|_\infty)^{N - 1}\|x - \tilde{x}\|_\infty, \] where \(\epsilon(\lambda,r)=\sup_{-r\leq z\leq r}|e^{\lambda z} - 1|\).