Machine Learning, is an issue that appears as early as 1959, when Arthur Samuel defined it as “a field of study that gives to computers feature of learning without being explicitly programmed.”
Generally, this topic is associated with the algorithms and techniques that allow in specific way self-learning and adaptation to the close environment.
From the Microsoft point of view, this concept, which deals with this area in the long term, is rather a Microsoft Azure service, which provide a new use of analytical services to be more effective in predicting of the new emerging applications and algorithms using large amounts of data in organizations and in the final making Forecasts about whatever the company likes. It is clear that with very large amounts of data there need to be created applications which are ideally become more adaptable rights to specific type of data, specific to the size of the databases that these data clump together and working with them as much effective as possible. Machine Learning in Microsoft’s submission called is called Azure ML: Predictive analyticsas a Service (PaaaS?) And the important thing is just that “AS A SERVICE” and certainly worth mentioning the word ANALYTICS.
With this service it is possible implementing quality prediction of all sorts. Whether it is a prediction of the amount of orders, goods sold, forecast the allocation of human resources, or demographic forecasts from a global perspective.
It is therefore possible to create an environment of “Machine Learning workspace” and then use the ML Studio web product, where you can create predictive models that adapt themselves, and it is also possible to interconnect numerous possible dates for these models.
It is possible to create separate models using the ML design studio, which allows you to import various data sources, integrate them and perform various simulations of such resources. For example, it is possible to import the MS Excel workbooks that are saved as. CVS file. Above these models it is possible to run scripts (in this case R Scripts) and use complex set of tools. The whole model behaves more like a workflow or activity diagram, if you will.
Excellent about this studio is the fact that if you right click on the action “execute the script”, you can visualize this result, which is shown in the following figure.
What more can I say.. Good luck with your modeling, I believe it will bring a lot of cost savings and a competitive advantage – to be one step ahead….
- https://en.wikipedia.org/wiki/Machine_learning - https://cs.wikipedia.org/wiki/Strojov%C3%A9_u%C4%8Den%C3%AD - https://chanel9.com