Porosity prediction of tight reservoir based on an integrated convolutional neural network and long short-term memory with differential evolution optimization

Key Points

• Porosity prediction accuracy significantly improves using an integrated convolutional neural network and long short-term memory.
• The model achieved an accuracy of 92.5% on test data, validating its robustness for geological assessments.
• Analysis utilized an integrated approach of neural networks and differential evolution optimization to enhance performance.
• This may enable better reservoir management strategies, though the findings need validation in varied geological settings.