SigmaWay Blog

SigmaWay Blog tries to aggregate original and third party content for the site users. It caters to articles on Process Improvement, Lean Six Sigma, Analytics, Market Intelligence, Training ,IT Services and industries which SigmaWay caters to

Working with Machine Learning

Artificial Intelligence, Machine Learning and Deep Learning are relatively newer technologies invading the fields of information technology, business etc. Though developers are walking towards this era, currently the number of experts is relatively less. The company often makes mistakes by starting up with the technologies instead of focusing on business needs. They often make mistakes by assigning out of domain work to some. For e.g. Hiring data scientists and asking them to build something interested from given database. Rather than a team must be formed of product managers, data engineers, data scientist and DevOps engineers.A team of four will be a kick start to improve our process and giving better results. Now everybody has an opportunity to improve the models, optimise the deployment and scale the business. 

Talking about ML, many projects fail due to complex structures. This could occur because of working on wrong problem, to having wrong data, failing to build a model or failing to deploy it correctly. Read more at: https://medium.com/@guyernest/the-flywheel-of-machine-learning-systems-50aa6d992382

Rate this blog entry:
1067 Hits
0 Comments

Building 21st Century Data Science Teams

A traditional data science department is comprised of Data Scientists, Data Engineers and Infrastructure Engineers. This model has a drawback that one role is always dependent on other and likely to criticize them for task failures because they didn't do their job well. These conflicts may reflect in the quality of final data product. So, what went wrong? You probably don't have big data. Jeff Magnusson (Director of Algorithms Platform at Stitch Fix) suggested a clever approach of forming a "High Functioning Data Science Department" which involves building an environment which allows autonomy, ownership, and focus for everyone involved yet at the same time clearly distinguishing the roles of Data Scientists and Data Engineers. Data scientist can't suddenly become talented engineers nor is that engineers will be ignorant of all business logic, the partnership is inherent to the success of this model. You can read more at: http://multithreaded.stitchfix.com/blog/2016/03/16/engineers-shouldnt-write-etl/

 

Rate this blog entry:
1429 Hits
0 Comments
Sign up for our newsletter

Follow us