Please enable JavaScript.
Coggle requires JavaScript to display documents.
Chapter 3: Intro to Predictive Modeling (Models, Induction, and Prediction…
Chapter 3: Intro to Predictive Modeling
Supervised Segmentation
how can we segment the population into groups that differ from each other with respect to some quantity of interest
Models, Induction, and Prediction
Model
Simplified representation of reality created to serve a purpose
It is simplified based on some assumption about what is and is not important for specific purpose
Ex. Map is a model of the world
In data science, a predictive model is a formula for estimating the unknown value of interest:
the target
Descriptive Modeling
primary purpose is not to estimate a value but to gain insight into the underlying phenomenon or process
Supervised learning is model creation where model describes a relationship between a set of selected variables (attributes or features) and a predefined variable called the target variable
Induction
the creation of models from data
the input data for the induction algorithm is called the training data
Supervised Segmentation
predictive model - estimates the value of some particular target variable
extracting patterns from data in a supervised manner - try to segment population into subgroups that have different value for the target variable
if segmentation is done using values of variables that will be known when the target is not known - then segments can be used to predict the value of target variable
Selecting Informative Attributes
Information gain
based on a purity measure called entropy
entropy tells us how impure an individual subset is
Supervised Segmentation with Tree-Structured Models
often used as predictive models
can predict classification by finding corresponding segment and using the class value ast the leaf
represent exactly the sort of supervised segmentation we often want
popular & easy to understand