According to the current situation, analysis become constantly optimize the business operations of the basic framework of the internal part of subjective feeling, now the data science is usually through a one-time effort, or temporary deployment to a production environment, thus the process of obtaining new insights and model, and the need for close monitoring and update regularly,

CodePudding user response:

Analysis reflects the focus resource intensive operation one-time victory and has continued to adapt to environmental changes "nourish the source of the difference between", in order to achieve this goal, enterprise organizations need to establish a cross-functional team, using the right software, and implement strict discipline, and let the data scientists, engineers, product managers and domain experts work together to create a continuous process to bring value for the enterprise,

The next step is to begin from balance spending and business development, so that you can to a certain amount of investment analysis, bridge the gap between the data science and IT project, if you don't use this prospective methods, enterprises may carry out some interesting analysis projects, these projects can run for a period of time, but will eventually decline, become less important, and unable to get progress, is the most frustrating, enterprise will eventually cannot be obtained from analysis of investment return on the level of implementation and deployment,

Analysis technology in the next step development will not only driven by data scientists, it needs for skills, practice and support technology, the applications of data analysis from the laboratory to the business, the analysis of the operation need to make decisions consciously, continuous integration, testing, deployment, monitoring, and adjust the analysis, so as to realize continuous improvement, the analysis work, no matter how complex, should not be regarded as a terminal project, and should be regarded as the entire operating an integral part of the framework,
From: the TechTarget big data at https://searchbi.techtarget.com.cn/microsite/4-4568/