ISU Theoretical and Applied Data Science (TADS) authors have had a paper receive an ACM SIGSOFT Distinguished Paper Award. Association for Computing Machinery Special Interest Group on Software Engineering (ACM SIGSOFT) encourages SIGSOFT-sponsored conferences to designate a number of accepted papers for ACM SIGSOFT Distinguished Paper Awards for the conference. The award is very competitive and only given to at most 10% of the papers accepted.
This paper co-authored by Ph.D. student Rangeet Pan and Dr. Hridesh Rajan entitled "On Decomposing a Deep Neural Network into Modules" was awarded at the ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering (ESEC/FSE) 2020. According to the paper's abstract, "Deep learning is being incorporated in many modern software systems. Deep learning approaches train a deep neural network (DNN) model using training examples, and then use the DNN model for prediction. While the structure of a DNN model as layers is observable, the model is treated in its entirety as a monolithic component. To change the logic implemented by the model, e.g. to add/remove logic that recognizes inputs belonging to a certain class, or to replace the logic with an alternative, the training examples need to be changed and the DNN needs to be retrained using the new set of examples. We argue that decomposing a DNN into DNN modules-akin to decomposing a monolithic software code into modules-can bring the benefits of modularity to deep learning. In this work, we develop a methodology for decomposing DNNs for multi-class problems into DNN modules. For four canonical problems, namely MNIST, EMNIST, FMNIST, and KMNIST, we demonstrate that such decomposition enables reuse of DNN modules to create different DNNs, enables replacement of one DNN module in a DNN with another without needing to retrain. The DNN models formed by composing DNN modules are at least as good as traditional monolithic DNNs in terms of test accuracy for our problems." So far deep neural networks, the machine learning models used for deep learning, are thought of as monolithic machine learning models. This work shows for the first time that it is possible to decompose a deep neural network into parts so that these parts can be reused to create another deep neural network and replaced to improve the machine learning model.
ESEC/FSE is an internationally renowned forum for researchers, practitioners, and educators to present and discuss the most recent innovations, trends, experiences, and challenges in the field of software engineering. Along with the International Conference on Software Engineering (ICSE), it is widely recognized as the top-2 conferences in software engineering. ESEC/FSE brings together experts from academia and industry to exchange the latest research results and trends as well as their practical application in all areas of software engineering.