List - Twelve unique characteristics of IoT based Predictive analytics/machine learning.

Over the holiday weekend, I have been working on a list of unique characteristics for IoT and Machine/Predictive analytics. 

It started off with 7 .. then 8 .. then 10 and now 12 characteristics

Here is my list. More soon in detailed blog posting

Qs is: Have I missed any? Each of these are very detailed in themselves and rapidly shifting

So, happy for any feedback - else twelve it is :) 

Twelve unique characteristics of IoT based Predictive analytics/machine learning. 

1) Time Series Data: Processing sensor data. 

2) Beyond sensing: Using Data for improving lives and businesses. 

3) Managing IoT Data. 

4) Rethinking the edges of the Enterprise: Supply Chain and CRM impact of IoT. 

5) Decisions at the 'Edge' 

6) Real time processing. 

7) Image processing. 

8) Managing Massive Geographic scale. 

9) Cloud and Virtualization. 

10) Integration with Hardware. 

11) Rethinking existing Machine Learning Algorithms  for the IoT world. 

12) IoT datasets and Training IoT algorithms. 

Views: 2118

Replies to This Discussion

PS - I am going to post a much more detailed post on each of these in a couple of days! But will be an evolving story which I shall develop here 

will be good to read a detailed post.

thanks Sandeep. Shall add them one by one else will be a very large post. However here is one of them which I posted as a main blog(ie not in the group)

'Datalogix like’ insights: A new #IoT #Retail #Beacon business mode...

I'm really curious to hear what you have to say about Integration with Hardware and Image Processing, Ajit.

indeed! ex see below from http://www.cognimem.com/

I think IoT will need a lot of predictive analytics at the edge

companies like Cisco and GE are calling it 'fog computing' 

but its been around for a while. 

there are some more good examples also the Image processing bit ..(ie not image processing per se - but through surveillance cameras etc) ex real time face recognition


more soon! 


I am wondering if you are planning to get into the dirty details of doing edge based data quality analysis based upon non-periodic time intervals between data faults?  This is where my interest lies and your image of the edge based analysis above seems to imply that you may be doing so (Anomaly Detection).   I can't wait to see what you have to say.  I am currently working on a project along those lines and any ideas/information that I can gather will help.

Are you planning on going into the RNN areas of prediction by maybe including TensorFlow and/or Keras?



© 2021   TechTarget, Inc.   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service