Copyright 2017 Project Botticelli Ltd unless noted otherwise

The information herein is for informational purposes only and represents the opinions and views of Project Botticelli and/or Rafal Lukawiecki. The material presented is not certain and may vary based on several factors. Project Botticelli Ltd and Microsoft make no warranties, express, implied or statutory, as to the information in this presentation. Copyright © 2017 Project Botticelli Ltd unless noted otherwise. Some slides contain quotations from copyrighted materials by other authors, as individually attributed, including under Creative Commons Attribution license, Wikimedia terms, Apache or a BSD license, or as already covered by Project Botticelli Ltd or Microsoft Copyright ownerships. All rights reserved. Microsoft, Windows, SQL Server, and other product names are or may be registered trademarks and/or trademarks in the U. S. and/or other countries. The information herein is for informational purposes only and represents the current view of Project Botticelli Ltd as of the date of this presentation. Because Project Botticelli & Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft and Project Botticelli cannot guarantee the accuracy of any information provided after the date of this presentation. Project Botticelli makes no warranties, express, implied or statutory, as to the information in this presentation. E&OE.

5 -day class Online projectbotticelli. com/courses projectbotticelli. com Dublin 16 Oct London 6 Nov Oslo 20 Nov Prague 4 Dec Stockholm 11 Dec Chicago Spring ‘ 18 Connect with me projectbotticelli. com/connect Copyright 2017 © Project Botticelli Ltd. All rights reserved. Study AI+DS+ML with me!

everything Accuracy Classification: precision, recall Regression: R 2, RMSE, Reliability Usefulness cross-validation standard deviation notched box plot, p-values

Copyright 2017 © Project Botticelli Ltd. All rights reserved.

your resources Select precision, Performance Select recall,

1. 2. standard deviation 1% 1 -4% Copyright 2017 © Project Botticelli Ltd. All

business testable 3. Iteratively

States the situation that we would like to achieve or test for, expected outcome Copyright 2016 © Project Botticelli Ltd. All rights reserved. States the situation as-is, status-quo, nothing new, “the pessimistic view”

not Type I error not Type II error Copyright 2016 © Project Botticelli Ltd.

control Run Evaluate Statistically Copyright 2017 © Project Botticelli Ltd. All rights reserved. Design

not Type I error wrong decision question data compatibility with your model

Probability chance as extreme as those observed Example: how unusual is getting 8 heads in 10 coin throws? 8 heads would be at least as extreme as getting 8, 9, or 10 heads (because 9 and 10 is more extreme than 8) p-value = p(8 out 10) + p(9 out 10) + p(10 out 10) = 0. 055

p-value incompatible observed data is with a single hypothesis significance level α make a decision different Incorrectly.

chance of no effect Long Shot of Toss Up Good Bet

p-value of a test hypothesis prior probability of the “Think Bayes” “The ASA’s Statement on p-Values, : Context, Process and Purpose”

odds true Bayes ratios confidence intervals conservative likely to be

enough conservative

best easy Copyright 2017 © Project Botticelli Ltd. All rights reserved. Understand the meaning

Connect with me projectbotticelli. com/connect @rafaldotnet 5 -day class projectbotticelli. com/courses Dublin, London, Oslo, Prague, Stockholm, Chicago

https: //myignite. microsoft. com/evaluations https: //aka. ms/ignite. mobileapp

Connect with me projectbotticelli. com/connect @rafaldotnet 5 -day class projectbotticelli. com/courses Dublin, London, Oslo, Prague, Stockholm, Chicago

Copyright 2017 © Project Botticelli Ltd. All rights reserved. The information herein is for informational purposes only and represents the opinions and views of Project Botticelli and/or Rafal Lukawiecki. The material presented is not certain and may vary based on several factors. Project Botticelli Ltd and Microsoft make no warranties, express, implied or statutory, as to the information in this presentation. Copyright © 2017 Project Botticelli Ltd unless noted otherwise. Some slides contain quotations from copyrighted materials by other authors, as individually attributed, including under Creative Commons Attribution license, Wikimedia terms, Apache or a BSD license, or as already covered by Project Botticelli Ltd or Microsoft Copyright ownerships. All rights reserved. Microsoft, Windows, SQL Server, and other product names are or may be registered trademarks and/or trademarks in the U. S. and/or other countries. The information herein is for informational purposes only and represents the current view of Project Botticelli Ltd as of the date of this presentation. Because Project Botticelli & Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft and Project Botticelli cannot guarantee the accuracy of any information provided after the date of this presentation. Project Botticelli makes no warranties, express, implied or statutory, as to the information in this presentation. E&OE.