Draggable machine learning workflow component scheduling method

Abstract

The invention discloses a draggable machine learning workflow component scheduling method, which comprises the following steps of S1, designing a component configuration template corresponding to a component category, and enabling other nodes except a pseudo node Base in the configuration template to be in one-to-one correspondence with tasks required by machine learning modeling; s2, acquiring tasks and a task sequence required by machine learning modeling included in the current machine learning workflow, and transmitting the task sequence and the configuration template as parameters; s3, dynamically loading and configuring template parameters according to the task sequence, and executing the machine learning workflow. According to the draggable machine learning workflow design thought, workflow modularization can be achieved, rapid reuse can be achieved, and creation of machine learning workflow tasks can be achieved through Web front-end page dragging, code layer and command line modes.

Description

technical field [0001] The invention relates to the technical field of machine learning workflow, in particular to a draggable method for scheduling machine learning workflow components. Background technique [0002] In recent years, with the rapid development of computer application technology, artificial intelligence, big data, and cloud computing have become the focus of attention in the IT field. As a machine learning (Machine Learning, ML) algorithm that makes computers "intelligent", it has achieved remarkable results in target recognition, target detection and other tasks, and has been successfully applied to financial transactions, commodity recommendations, traffic prediction and other fields. When using a machine learning algorithm to train a model, in order to avoid excessive time-consuming processes such as raw data collection, data cleaning, missing value processing, feature extraction, sample generation, and model evaluation, it is usually necessary to build a ...

AI Technical Summary

Problems solved by technology

This idea of ​​machine learning workflow construction is opposed to the way of building and testing machine learning algorithms through the code layer. The corresponding workflow scheduling method is also limited and cannot adapt to the way of building based on the code layer.

Images