Utilizing a Long Short-Term Memory Algorithm Modified by Dwarf Mongoose Optimization to Predict Thermal Expansion of Cu-Al2O3 Nanocomposites

This paper presents a machine learning model to predict the effect of Al2O3 nanoparticle content on coefficient thermal expansion in Cu-Al2O3 nanocomposites prepared using an situ chemical technique. The developed is modification Long Short-Term Memory (LSTM) dwarf mongoose optimization (DMO), which mimics behavior DMO find its food for predicting composite. swarm consists three groups, namely alpha group, scouts, and babysitters. Each group has own capture food. preparation nanocomposite was performed aluminum nitrate that added solution containing scattered copper nitrate. After that, powders CuO were obtained, leftover liquid removed treatment at 850 °C 1 h. consolidated compaction sintering processes. impact contents properties investigated. results showed Thermal Expansion Coefficient (TEC) decreases with increasing due increased precipitation nanoparticles grain boundaries Cu matrix. Moreover, good interfacial bonding between may participate this decrease TEC. proposed able TEC all produced composites different tested temperatures very accuracy, reaching 99%.