Title

DeeperCoder: Code Generation Using Machine Learning

Publication Date

1-1-2020

Document Type

Conference Proceeding

Department

Applied Data Science

Publication Title

2020 10th Annual Computing and Communication Workshop and Conference, CCWC 2020

DOI

10.1109/CCWC47524.2020.9031149

First Page

194

Last Page

199

Abstract

In this paper, we present a program generation system based on input and output specification. The system is developed based on a programming-by-example technique used in program synthesis. The system can generate computer programs that satisfies user requirements based on inputs and outputs. We created a simple Domain Specific Language (DSL) which will be used in program synthesis. We trained our neural network with a large set of input space and store corresponding sample training programs. To get the final output which satisfies all the user specifications, we used inductive program synthesis and machine learning. We also experimented with different deep learning models to obtain the desired results with reduced number of steps and execution time. Finally, we show three layers of neural networks with LeakyReLU achieves the best performance when compared to other approaches.

Keywords

Deep Learning, DeepCoder, Domain Specific Language, Neural Networks, Program Synthesis, Programming Languages

Share

COinS