Great Final Takeaways from 2017 for the Financial Markets

Great short blog post by Charles Rotblut in the tweet below.

These two ideas really resonate with me:

  1. Staying out of the market because you didn’t like President Trump or President Obama has been very costly
  2. If didn’t buy bitcoin early in 2017, question your ability to predict what it will do now or in the future

Using Azure ML to Predict Customer Credit

Simple video example of how Azure ML (Machine Learning) classification model can be used to predict whether a bank customer is likely to have good or bad credit.

I am just starting to learn in this area, but I believe the key is a good training input data set.

At work, perhaps this can be used to analyze the first 10 to 15 minutes of a sales call to gauge intent or likely-hood of a purchase.

This is courtesy of Microsoft’s Edx.org Data Science Series.

Here is a link to the portfolio where I ran the same test:

https://gallery.cortanaintelligence.com/Experiment/Bank-Credit-Experiment-created-on-12-16-2017

Simulate Histogram of Company Profits

Microsoft EdX Course on Data Science Essentials, around Simulation of Customer profits, given formulas and estimates for Number of Customers, Profits Per Cup, and Tips!

Imagine how you can scale that up from a Coffee/Lemonade stand business to a major corporation!

These are great summary notes in case you need a refresher on basics around distributions, hypothesis, types of tests (Single Sample T and Z Tests, Two Sample Tests), and confidence intervals:

https://github.com/MicrosoftLearning/Data-Science-Essentials/raw/master/Handouts/Simulations.pdf

https://github.com/MicrosoftLearning/Data-Science-Essentials/raw/master/Handouts/Hypothesis%20Testing.pdf

Cleaning an example data set: https://gallery.cortanaintelligence.com/Experiment/Autos-Experiment-created-on-12-17-2017