RLML: A Domain-specific Modelling Language for Reinforcement Learning

In recent years, machine learning technologies have gained intense popularity and are being used in a wide range of domains. However, due to the complexity associated with machine learning algorithms, it is a challenge to make it user-friendly, easy to understand and implement. Machine learning applications are especially challenging for users who do not have proficiency in this area. In this work, we use model-driven engineering (MDE) methods and tools for developing a domain-specific modelling language (DSML) to contribute towards providing a solution for this problem. We targeted reinforcement learning domain from machine learning technologies, and evaluated the proposed language with multiple applications. We built a domain-specific modelling environment to support our reinforcement learning modelling language (RLML). The tool supports syntax-directed editing, constraint checking, and automatic generation of code from RLML models. With our proposed approach, we were able to move away from the complexity of implementing machine learning algorithms in general purpose languages and offer abstraction and simplicity for non-experts, which are a few of the characteristics and benefits of modelling languages.