This paper introduces a spectral graph wavelet transform for functions defined on the vertices of an arbitrary finite weighted graph. The transform is based on the spectral decomposition of the graph Laplacian, which provides an analogue of the Fourier domain for graph signals. The wavelet operators are defined using a generating kernel $ g $ and a scale parameter $ t $, with the scaled wavelet operator $ T_g^t = g(t\mathcal{L}) $. The wavelets are obtained by applying these operators to an indicator function, and the transform is invertible under an admissibility condition on $ g $. The paper explores the localization properties of the wavelets in the limit of fine scales and presents a fast Chebyshev polynomial approximation method for computing the transform without diagonalizing the Laplacian. The transform is shown to be invertible and forms a frame when the scale parameter is sampled at a finite number of values. The paper also discusses applications of the transform to various problem domains, including signal processing, pattern recognition, and data analysis on graphs. The spectral graph wavelet transform is shown to be analogous to the classical wavelet transform, with the wavelet coefficients representing inner products with the graph wavelets. The paper concludes with a detailed analysis of the transform's properties, including its invertibility, localization, and frame bounds.This paper introduces a spectral graph wavelet transform for functions defined on the vertices of an arbitrary finite weighted graph. The transform is based on the spectral decomposition of the graph Laplacian, which provides an analogue of the Fourier domain for graph signals. The wavelet operators are defined using a generating kernel $ g $ and a scale parameter $ t $, with the scaled wavelet operator $ T_g^t = g(t\mathcal{L}) $. The wavelets are obtained by applying these operators to an indicator function, and the transform is invertible under an admissibility condition on $ g $. The paper explores the localization properties of the wavelets in the limit of fine scales and presents a fast Chebyshev polynomial approximation method for computing the transform without diagonalizing the Laplacian. The transform is shown to be invertible and forms a frame when the scale parameter is sampled at a finite number of values. The paper also discusses applications of the transform to various problem domains, including signal processing, pattern recognition, and data analysis on graphs. The spectral graph wavelet transform is shown to be analogous to the classical wavelet transform, with the wavelet coefficients representing inner products with the graph wavelets. The paper concludes with a detailed analysis of the transform's properties, including its invertibility, localization, and frame bounds.