ICSE 2023 (series) / Artifact Evaluation /
Artifact for the paper - "Decomposing a Recurrent Neural Network into Modules for Enabling Reusability and Replacement"
The artifact is intended for a “functional/available” badge for ICSE 2023 accepted paper: “Decomposing a Recurrent Neural Network into Modules for Enabling Reusability and Replacement.” This research aims to enable the fine-grained reuse of RNN artifacts, a class of deep-learning algorithms. To evaluate this artifact, a basic understanding of deep learning, particularly RNNs, and some familiarity with the Python deep-learning ecosystem are needed (Tensorflow/Keras). A guide to locating the detailed instructions is provided to reproduce the results in PDF attached