Machine Learning

Just out of curiosity how to determine the number of tags required to perform a simple machine learning application? Is there any such a formula exist to build the algorithm? I do understand this is a broad question and of course depends on many factors…
Thanks… Sam

1 Like

Minimum number is one tag, with some history that you can learn from. But it does indeed depend on many factors. :slight_smile:

There are a lot of formulas, and that’s where it gets tricky. What does your data look like? What are you trying to predict. How noisy is your data? Do you have missing data? And so much more.

I’d suggest doing some background reading on ML. Here’s a couple of sites I thought were good fast (but not complete) intros: https://ml.berkeley.edu/blog/2016/11/06/tutorial-1/ and https://www.digitalocean.com/community/tutorials/an-introduction-to-machine-learning . They won’t give you enough to write your own ML application, but they’ll give you a taste of what can be done and some of the questions you need to answer.

Shameless plug: I’d also suggest watching the ICC 2017 session that Kevin M. and I did. (https://icc.inductiveautomation.com/archive#) That will also give you a great over view of types of machine learning and what you can do. Again, you won’t have enough to start your project, but you’ll know some of the questions to ask

If you want to really get into it, at a minimum I’d suggest Andrew Ng’s free course: https://www.coursera.org/learn/machine-learning That will definitely give you enough to write your own (beginning) application.

2 Likes

A demo ML project on Ignition would help us a lot to kick start. We can also proudly claim, “the first ML powered SCADA in the world”, which can become a fantastic marketing pitch for Ignition. Give us a weapon to show the decision makers, the difference between Ignition and others.

The best sales pitch: “Show the buyer, how he can add value and make money with your product. He will become your salesman !!”.

There’s a demo project attached to the video. :slight_smile:

I’ll also drop a teaser that we’ll be adding something to 7.9 that will make your ML and analytics projects easier. Now you all have to wait in agony to find out what it is. :wink:

5 Likes

Hi
I am Ankit Tiwari
I have some basic information

Only one tag required~>
6 Steps To Write Any Machine Learning Algorithm From Scratch: Perceptron Case Study**

  1. Get a basic understanding of the algorithm.
  2. Find some different learning sources.
  3. Break the algorithm into chunks.
  4. Start with a simple example.
  5. Validate with a trusted implementation.
  6. Write up your process.