Please enable JavaScript.
Coggle requires JavaScript to display documents.
Chapters 2-7 (Chapter 2 (Automated Machine Learning (supervised and…
Chapters 2-7
Chapter 2
examples of ML
Alexa
Google
using AI for automated cars
Siri
purposes of ML
automate hundreds of
millions of jobs that were
considered too complex
for machines to ever
take over
ML tools are
integral for knowing
the customer,
streamlining
existing operations,
and managing risk
and compliance
Automated Machine Learning
can be compared to
human learning
supervised and unsupervised
supervised = scientists select what
they want the
machine to learn
enables computers
to learn to solve a
problem by
generalizing from
examples (historical
data), avoiding the
need to explicitly
program the
solution
must have data for the machine
not automatic machine learning
improves work performance
general platforms
8 criteria for AutoML excellence
accuracy, productivity, ease of use, understanding and learning, resource availability, process transparency, generalizable across contexts, recommend actions
machine learning life cycle: 5 stages
Chapter 3-4
chapter 3
specify a business problem
opportunity
examples
which customers are satisfied
which customers are likely to purchase a product
why customers do/do not purchase the company’s products
problem statements
impact on bottom line, language of business, actions from results
chapter 4
acquire subject matter expertise
setting realistic expectations for model performance
data collection suggestions
SME
Chapter 5-6
chapter 5
decide on a unit of analysis
UOA: 4 Ws of analysis
think about what you are trying to analyze
chapter 6
define prediction target
prediction target: behavior of a thing we need to know about in the future
ex: predicting if a loan is bad
without a target, there is simply no way for humans or machines to learn what associations drive an outcome
classification: predicts the category to which a new case belongs
regression: predict the target numeric values
Chapter 7
Success, Risk, and Continuation
success
identify success criteria
risks
difficult to calculate
model risks, ethical risks, cultural risks, and environmental risks
decide whether to continue
weigh risks against rewards