Project Requirements

The training management system must support multiple advanced platforms, multiple training environments, and management and research needs—for example big-data and AI needs on the platform; day-to-day deep-learning training management of hardware and software configuration and student and teacher authorization in the environment; integrated management of courses, experiments, grades, users, resources and experimental environments; and subsequent seamless integrated management of the big-data training platform and integration with the research platform. The aim is to cultivate senior specialists with a sound grounding in mathematics and economics, mastery of basic statistical theory and methods, and skilled use of computer technology to analyze data, who can undertake statistical surveys, data analysis and other statistical practice in enterprises and public institutions, and financial-investment analysis, research and application in banks, securities firms, investment companies, insurance companies and other financial institutions.

Solution

The project uses Sugon’s self-developed big-data and AI training platform, aimed at university talent development in big data and AI. Drawing on rich experience in education as well as servers, big data and AI, Sugon has fully built the big-data and AI training-platform product. The Sugon XData-EDU training platform provides users with a complete one-stop big-data and AI teaching and training solution in an integrated hardware-and-software form. It mainly includes a unified teaching-management system, experimental environments, development tools, a container-scheduling system, a user-management system, an O&M management system, a curriculum system and an online examination system. The system uses advanced, flexible, lightweight container technology to provide big-data and AI experimental environments. Users can have exclusive container clusters and configure experimental resources on demand, solving resource contention and data-security problems in multi-user environments.

To meet different teaching and experiment needs of teachers and students, the system includes built-in Java, Python, R, Jupyter and other integrated development environments for big-data and AI experiments. In the integrated development environments, interactive debugging and running of code are supported. Users submit jobs directly to the back end, and the runtime can execute in a distributed way.

The training platform provides an enterprise-class container experimental environment; uses mainstream container technology to provide a lightweight, flexible and easy-to-use cluster environment; supports rapid cluster creation and uses Docker to build flexible experimental environments; and supports management and scheduling of more than a hundred container experimental clusters at once. Each student can have an exclusive container experimental environment, without affecting others.

Customer Benefits

The project meets the construction goals of the statistics big-data laboratory at North China University of Technology, including textbooks, experiment guides, experimental data and reference code. It can complete cultivation of interdisciplinary, application-oriented senior specialists for statistical information management, statistical analysis and forecasting, market research and consulting, or securities and futures trading, and brings the following benefits:

• A big-data research cluster and a complete teaching system oriented to the statistics big-data center, meeting integrated teaching and research;

• An experiment and development environment based on a web-notebook model, supporting learning and experiments anytime, anywhere;

• Online expansion of function modules, meeting future business expansion;

• Supporting training courses, providing users with localized, fast on-site technical support.