Implementation of Microsoft Azure for API Generation

Azure Machine Learning can be used for any type of machine learning, from classical ml to in-depth, supervised, and transfer learning. Whether we choose to write code in Python or R with the SDK or work with non-coded / low-code studio options, we can create, train, and track machine learning and in-depth learning models in the Azure Machine Learning Workspace. Azure Machine Learning Studio is a website on Azure Machine Learning for low-cost options and model training codes, shipping, and asset management. The studio merges with the Azure Machine Learning SDK for a seamless experience. In-studio we applied two-class decision jungle machine learning algorithms. The corresponding algorithm allows tree branches to come together, the directed acyclic graphs (DAG) decision often has lower memory and better performance to do than a decision tree, even if it costs a certain amount of training longer. Decision jungles are non-parasitic species that can govern the boundaries of offline decisions. They make the selection and classification of features integrated and are able to be strong when there are sound features. The accuracy achieved was approximately 99% and to be precise the model is not overfitting. The model was converted into a forecasting experiment which helps to deploy the model using web services and thus it can be called by REST API and HTTP call. The below figure shows the demonstration of the entire web service on azure ML studio