Machine Learning · Clief Notes

Machine Learning

This one goes back a few years. I have an excel spreadsheet with about 30 columns that represent submittals that were individually reviewed over the last 5 years. Some were told yes; some were told no. We have about 10,000 examples. I want to see if Machine Learning can predict future submissions

I have an ICM that builds other ICMs for me. This one is essentially controlled brainstorming. But these stages will help me stay organized.

I want to see how machine learning does in terms of accuracy and solution. In about 30 minutes, I am going to type, 'what kind of machine learning is best to solve this problem'. It is a magical time that we have a tool that can guide us through examples like this - we can go from concepts to concrete solutions without a heavy textbook next to us.

A few decades ago, linear regression was familiar to me. Backpropagation is very similar math to calculating thermodynamic properties during combustion. But I was raising kids while Machine Learning was king. So this is my chance to fill in the gaps - just a very small bit, I know.

5 comments

Machine Learning