Application of Generalised Regression Neural Networks in Lossless Data Compression

Neural networks are a popular technology that exploits massive parallelism and distributed storage and processing for speed and error tolerance. Most neural networks tend to rely on linear, step or sigmoidal activation functions for decision making. The generalised regression neural network (GRNN) is a radial basis network (RBN) which uses the Gaussian activation function in its processing element (PE). This paper proposes the use of the GRNN for lossless data compression. It is applied in the first stage of the lossless twostage predictor-encoder scheme. Three different approaches using the GRNN are proposed. Batch training with different block sizes is applied to each approach. Two popular encoders, namely arithmetic coding and Huffman coding, are used in the second stage. The performance of the proposed singleand two-stage schemes are evaluated in terms of the compression ratios achieved for telemetry data test files of different sizes and distributions. It is shown that the compression performance of the GRNN schemes is better than existing implementations using the finite impulse response (FIR) and adaptive normalised least mean squares (NLMS) filters, as well as an implementation using a recurrent neural network. Key-Words: Lossless data compression, neural network, two-stage, predictor, encoder, radial basis.