<p>This paper proposes a novel lossless audio compression method that enhances compression efficiency by integrating an Artificial Neural Network (ANN)-based delta addition model with a specialized clustering algorithm for magnitude components. The proposed Adaptive Magnitude Clustering (AMC) algorithm employs a dynamic step-size rule to partition the magnitudes into multiple clusters. By optimizing the trade-off between system constraints and mean square error (MSE), the AMC effectively manages the number of cluster formations. An ANN based delta addition method is employed to further reduce the mean square error (MSE) by fine-tuning the predicted values to more closely match the actual values. The effectiveness of the proposed lossless audio compressor is evaluated and compared with state-of-the-art lossless audio coding techniques. Experimental results indicate that the proposed method outperforms existing approaches in terms of compression efficiency while preserving lossless quality.</p>