Framework for Platform Independent Machine Learning (ML) Model Execution

This research study addresses the difficulty often encountered when deploying Machine Learning (ML) models into real-world applications. More than just building and training the model, deploying the ML model also includes making the model available for use in production environments, such as web applications, mobile apps, or Internet of Things (IoT) devices. The proposed solution suggests a novel approach to alleviate these deployment challenges. Traditionally, ML models are trained and deployed together in the same environment. However, the paper suggests decoupling this process. Instead of deploying the entire model, the focus shifts to splitting the execution of model related tasks (Training and Prediction) in separate environments. By adopting this approach, developers can take advantage of cloud resources for training while ensuring predictions can be made efficiently and quickly on different platforms. It also allows for easier updates and maintenance of ML models.