中文

Data Science Platform

A platform that offers highly automated and user-friendly machine learning algorithms, enabling users to rapidly build and deploy high-precision machine learning models, enhancing data mining productivity, and helping enterprises transition from the BI era to the AI era.

Product FeaturesFunction OverviewApplication Scenarios

Product Features

Highly Automated

Construct a comprehensive data mining workflow, covering data cleaning, feature engineering, machine learning algorithms, and online prediction.

Simple and Flexible

Hundreds of machine learning algorithms are encapsulated, allowing model training through visual drag-and-drop operations. Custom components can also be built according to business scenarios.

Automatic Feature Generation

By configuration, high-dimensional features can be systematically and automatically generated, significantly reducing the workload of manual feature creation.

Automated Learning

The system automates preprocessing, feature engineering, algorithm selection, and model training parameter tuning, lowering the cost of learning and usage.

Model Visualization

Provide intuitive and easy-to-understand model visualization features, making models more transparent and controllable, and facilitating analysis and optimization of models.

Diverse Services

Support both offline model prediction and online model prediction services, seamlessly connecting the entire machine learning workflow.

Function Overview

Visual Model Construction

Provide a WEB IDE visual development interface that enables data mining and data analysis functions through drag-and-drop without the need for coding.

Application Scenarios

One-Stop Algorithm Modeling

Provide a one-stop service for machine learning algorithms, covering data processing, feature engineering, model training, service deployment, and prediction.

Interactive Model Development

Conduct data cleaning, statistical analysis, data visualization, and machine learning model construction in a one-stop, interactive manner based on Notebooks.

AutoML Automated Modeling

Without requiring machine learning skills or manual intervention, the entire process from feature construction and feature combination to algorithm selection and parameter tuning is completed automatically, lowering the barrier to entry for AI applications.