Feature Engineering and Selection: A Practical Approach for Predictive ModelsCRC Press, 25.07.2019 - 310 Seiten The process of developing predictive models includes many stages. Most resources focus on the modeling algorithms but neglect other critical aspects of the modeling process. This book describes techniques for finding the best representations of predictors for modeling and for nding the best subset of predictors for improving model performance. A variety of example data sets are used to illustrate the techniques along with R programs for reproducing the results. |
Inhalt
| 1 | |
Illustrative Example Predicting Risk of Ischemic Stroke | 21 |
A Review of the Predictive Modeling Process | 35 |
Exploratory Visualizations | 65 |
Encoding Categorical Predictors | 93 |
Engineering Numeric Predictors | 121 |
Detecting Interaction Effects | 157 |
Handling Missing Data | 187 |
Working with Profile Data | 205 |
Feature Selection Overview | 227 |
Greedy Search Methods | 241 |
Global Search Methods | 257 |
Bibliography | 283 |
| 295 | |
Andere Ausgaben - Alle anzeigen
Feature Engineering and Selection: A Practical Approach for Predictive Models Max Kuhn,Kjell Johnson Eingeschränkte Leseprobe - 2019 |
Feature Engineering and Selection: A Practical Approach for Predictive Models Max Kuhn,KJELL. JOHNSON Keine Leseprobe verfügbar - 2021 |
Feature Engineering and Selection: A Practical Approach for Predictive Models Max Kuhn,Kjell Johnson Keine Leseprobe verfügbar - 2019 |
Häufige Begriffe und Wortgruppen
additional algorithm amount analysis applied approach assessment associated average better bioreactor categorical chapter characteristics combination complete component computed consider contain correlation created cross-validation data set determine discussed distribution dummy variables effect encoding engineering estimate evaluated example factors feature selection Figure filter function identify illustrate important improve imputation included increases indicates individual initial interaction iterations less Line linear mean measure methods metric missing missing values occur optimal original outcome overall p-value parameter patterns performance plot points positive possible potential predictive predictors probability problem procedure profiles random reason reduce regression relationship represented resampling response ridership samples scores shown shows similar simple specific split stations statistic STEM steps subset summarized Table techniques training set transformation trees trend tuning understand unit values variables variation visualization zero