New hybrid model accurately predicts software failures, saving costs and improving quality.
Software reliability is crucial for ensuring quality in software development. A new model combining two popular prediction methods has been developed to accurately predict software failures. By using the Markov method based on residual errors, the hybrid approach was tested on a U.S. Navy system and showed high accuracy in predicting software reliability.