A huge litany of business literature now exists on the scope and role of Machine Learning (ML) both for the present and future. While some doomsayers predict ML will take away all human jobs, others say machines and robots will simply supplement human work to help us achieve tasks more efficiently. However, more than being game changers, ML algorithms will have to steer clear of becoming the new PowerPoint presentations. No one intends to have boring presentations, yet many end up doing so. These end up sucking the life out of otherwise engaging meetings or discussions. Similarly, inadvertently, a lot of programmers will end up developing algorithms which will draw up false conclusions. This will be true for both Active and Passive MLs. In the former, a human programmer is constantly working to improvise on the results, while the latter is self-sustaining. To avoid this happening en masse, certain steps have been suggested by experts. The organization concerned must start off by writing a Declaration of Machine Intelligence which will outline the tasks the management expects the algorithms to perform. Proper transparency measures must be set up with grounds for reviews, validation and verification. As large corporations perform data warehousing tasks with these huge chunks of data, a lot of it will become junk. Thus, they need to be periodically reviewed by a panel. Finally, a trade-off road map must be curated so that when the algorithm fails, there is enough scope to improvise.


Uploaded Date:06 February 2018

SKYLINE Knowledge Centre

Phone: 9971700059,9810877385
© 2017 SKYLINE. All right Reserved.