Advanced Analytics

Note: The modules in this chapter are product advanced modules and must be purchased separately: Y-Advanced Analytics. From Version 9.0, you need to purchase Y-AI.

With the rapid development of computer technology and network technology, various enterprises have accumulated a lot of business data. These data, not only reflect the enterprise's operating conditions, but also have huge potential commercial values. Faced with mass data, the enterprise needs a powerful data processing tool which can accurately and efficiently detect the data, reflect the data fact, reveal data regularity, and provide reliable information for enterprise management decision. However, traditional data analysis is limited to reflecting the past and present business conditions of the enterprise. The manager cannot draw the conclusion directly from the dashboard presentation. Enterprise data can not play real value.

The advanced analytics function of Yonghong Z-Suite integrates complex statistical algorithms and machine learning technologies which can explore the relationships, patterns and trends of potential value, build data models and make forecast analysis from mass data. Help the enterprise to understand their own problems in a timely manner thus to discover market opportunities and make scientific management decisions.

Yonghong advanced analytics, as an important component of Yonghong one-stop data analytics platform, is upgraded since version 7.5 to achieve procedural operation and analytics. As an independent functional module, deep analysis has a new operating interface which has realized more professional machine learning functions. With the help of the intuitive workflow of advanced analytics, you can split data sets into test sets and training sets. Select the feature column and target column. Select an analysis algorithm, build an algorithm model, train the model and get model parameters. Use the test set to score the model and adjust the parameters/features to make the models more accurate and able to be applied to data set and visual dashboard, thus obtaining the predictive analytics results. Of course, Yonghong advanced analytics still contains traditional rapid analysis. Rapid analysis is to create analysis algorithms on data sets or chart components quickly, and rapidly bind to components to achieve visual results.

The procedural advanced analytics function of Yonghong Z-Suite provides classic statistical methods, such as Logistic Regression, K-Means Clustering, Time Series Analysis, Association Rules and Decision Tree to meet frequently used analysis scenarios. Relying on simple intuitive operations, you can establish models easily to complete prediction analysis, or call R package functions and customize analysis algorithms as R model users to realize more abundant and advanced statistical analysis and predictive analysis functions in R. New Python script node has been added since version 8.5.1 to adapt to more complex analysis scenarios. The 9.0 version adds plug-ins, which replace the five types of nodes of data transformation, statistical analysis, algorithm, model integration and verification and evaluation with plug-ins, and supports user-defined plug-ins, so that the number of advanced analysis operators is not limited to the built-in nodes, more types of plug-ins can be integrated continuously.

ML1

Note: 1. Users who only purchased Y-Advanced Analytics cannot use the new plug-in functions and WEB service functions of version 9.0.

2. Users who only purchased Y-AI cannot use the functions of the old version, and the old operation folder is not displayed.

3. For users who have purchased two modules at the same time, the operation catalog shows the new plug-in function of 9.0 and the WEB service function, and the old version of the operator can also be used. These operators are collected in the old operation folder of the operation tree, by default Hidden, can be displayed by modifying the bi parameter old.operator.folder.show=true.