Enterprises have accumulated a large quantity of data in the process of informationization, including its service, operation, finance and so on. As the increase of data dimensions and the degree of correlation, the traditional analytical method of BI can no longer fully dig out the value of data. The secondly emerging AI provides for data with a new application model, which can reveal deep relations in the data of high dimension, and precisely make classification and prediction in order to be the foundation of the decisions of the enterprise. Meanwhile, the builder of AI's application to the enterprise has new demands.
First and foremost, AI requires that the builder has the ability to reflect his business into the data. On one hand, the builder needs to be familiar with the process, contents and requirements of the business. On the other hand, he needs to know the currently accessible data resources and connect the requirements of business with data resources together.
Secondly, AI requires that the builder has the ability to develop the data modeling and algorithm modeling. The builder has to possess professional knowledge on such realms like statistics, machine learning and deep learning etc. Besides, he can develop the data modeling and algorithm modeling according to the requirements of business scenarios. In terms of these models, he can do practice, examinations and optimization so that they can satisfy the requirements of business application.
Furthermore, the application builders of enterprise need to transform the traditional application-building method, understand the application’s content and working mode at the AI level. Finally apply a new-type application for regression、classification、prediction、association discovery and other types of scenes.
Thus, a solution which can help enterprises to use AI model and algorithm easily and quickly, has become an urgent requirement.
The solution we provided includes mainstream machine learning and deep learning algorithm, supports clustering and dimensionality reduction、classification analysis、regression prediction、association detection、statistical analysis、data visualization and other intelligent analysis scenarios, meanwhile, provides mainstream algorithm and modeling facing algorithm process training.
We offer interactive and WYSIWYG analyzing and process modeling interface. It supports users to set up testbed for AI analysis. By trying different algorithm intuitively, we establish data analysis process and model, and further train、test、verification and update the process.