Learns Gradient Boosted Trees with the objective of classification. The algorithm uses very shallow regression trees and a special form of boosting to build an ensemble of trees. The implementation follows the algorithm in section 4.6 of the paper "Greedy Function Approximation: A Gradient Boosting Machine" by Jerome H. Friedman (1999). For more information you can also take a look at this.
The used base learner for this ensemble method is a simple regression tree as it is used in the Tree Ensemble , Random Forest and Simple Regression Tree nodes. Per default a tree is build using binary splits for numeric and nominal attributes (the later can be changed to multiway splits). The built-in missing value handling tries to find the best direction for missing values to go to by testing each possible direction and selecting the one yielding the best result (i.e. largest gain).
Sampling
This node allows to perform row sampling (bagging) and attribute sampling (attribute bagging) similar to the random forest* and tree ensemble nodes. If sampling is used this is usually referred to as Stochastic Gradient Boosted Trees. The respective settings can be found in the Advanced Options tab.
(*) RANDOM FORESTS is a registered trademark of Minitab, LLC and is used with Minitab’s permission.
Options
Advanced Options
Popular Predecessors
Scorer21 %Number To String9 %- Color Manager5 %
File Reader4 %- Decision Tree Predictor4 %
Joiner3 %- Row Sampling3 %
Excel Reader (XLS)3 %Row Filter (deprecated)2 %- Low Variance Filter1 %
Column Appender1 %- Correlation Filter1 %
Row Splitter (deprecated)1 %- Bootstrap Sampling1 %
String To Number1 %- ROC Curve (legacy)1 %
- Boosting Learner Loop Start1 %
- k-Means< 1 %
PCA< 1 %- Shuffle< 1 %
- One to Many< 1 %
- Rule Engine< 1 %
- Math Formula< 1 %
- Feature Selection Filter< 1 %
- Cluster Assigner< 1 %
- Hierarchical Clustering< 1 %
Linear Correlation< 1 %- GroupBy< 1 %
- Missing Value< 1 %
Column Rename< 1 %- Rule-based Row Filter< 1 %
Number To String (PMML)< 1 %- Google Sheets Reader< 1 %
- Table Creator< 1 %
Gradient Boosted Trees Predictor< 1 %Random Forest Predictor< 1 %- Category to Number< 1 %
- Denormalizer< 1 %
- Sorter< 1 %
- Data Generator< 1 %
Excel Reader (XLS)< 1 %Extract Date&Time Fields< 1 %Gradient Boosted Trees Predictor< 1 %- Simple Regression Tree Predictor< 1 %
- Target Shuffling< 1 %
Table Reader< 1 %Loop End< 1 %Merge Variables< 1 %XLS Reader< 1 %- Linear Regression Learner< 1 %
- Double to Integer< 1 %
Table Row to Variable< 1 %- ARFF Reader< 1 %
PCA Apply< 1 %- SMOTE< 1 %
Random Forest Learner< 1 %- Domain Calculator< 1 %
- Missing Value (Apply)< 1 %
- Normalizer (PMML)< 1 %
CSV Reader< 1 %Backward Feature Elimination Start (1:1)< 1 %- Chunk Loop Start< 1 %
- Group Loop Start< 1 %
- X-Aggregator< 1 %
- Fuzzy c-Means< 1 %
Concatenate (Optional in)< 1 %- Filter Apply< 1 %
- Normalizer< 1 %
- RowID< 1 %
- String Manipulation< 1 %
- Table Transposer< 1 %
- Statistics< 1 %
- Math Formula (Variable)< 1 %
- Integer Input (legacy)< 1 %
Interval Loop Start< 1 %Tree Ensemble Learner< 1 %- Tree Ensemble Learner< 1 %
- Number Rounder< 1 %
- Column Splitter< 1 %
- Decision Tree to Ruleset< 1 %
- Numeric Outliers< 1 %
- Box Plot (legacy)< 1 %
- Parallel Chunk Start< 1 %
- Boosting Learner Loop End< 1 %
- String Input (legacy)< 1 %
- Date&Time-based Row Filter< 1 %
- Try (Variable Ports)< 1 %
Database Reader< 1 %Backward Feature Elimination Start (2:2)< 1 %- Feature Selection Loop Start (1:1)< 1 %
- X-Partitioner< 1 %
- Decision Tree Learner< 1 %
PCA Compute< 1 %- Logistic Regression Predictor< 1 %
- SVM Predictor< 1 %
Gradient Boosted Trees Learner< 1 %- Concatenate< 1 %
- Expand Bit Vector< 1 %
- Number to Category (Apply)< 1 %
- Cross Joiner< 1 %
- Equal Size Sampling< 1 %
- Column Filter< 1 %
- Reference Column Filter< 1 %
- Nominal Value Row Filter< 1 %
- Reference Row Filter< 1 %
- Reference Row Splitter< 1 %
- Partitioning< 1 %
- Ungroup< 1 %
- Empty Table Switch< 1 %
- IF Switch< 1 %
CASE Switch Data (Start)< 1 %- SDF Reader< 1 %
- Empty Table Creator< 1 %
- PMML to Cell< 1 %
- SAS7BDAT Reader< 1 %
- Parameter Optimization Loop Start< 1 %
- R Snippet< 1 %
- Date&Time Difference< 1 %
- Information Gain Calculator< 1 %
- Outlier Removal< 1 %
- Table to String< 1 %
Auto Active Learn Loop End< 1 %- Strings to Binary Objects< 1 %
- Try (Data Ports)< 1 %
Table Row To Variable Loop Start< 1 %- Java Snippet< 1 %
- Feature Selection Loop Start (2:2)< 1 %
Naive Bayes Predictor< 1 %- Naive Bayes Predictor< 1 %
PCA Inversion< 1 %- Polynomial Regression Learner< 1 %
Numeric Scorer< 1 %- SVM Learner< 1 %
PMML Gradient Boosted Trees Predictor< 1 %- Random Forest Learner< 1 %
- Random Forest Learner (Regression)< 1 %
- Auto-Binner< 1 %
- Numeric Binner< 1 %
Cell Splitter< 1 %- Lag Column< 1 %
- Edit Nominal Domain< 1 %
- Edit Numeric Domain< 1 %
- Extract Missing Value Cause< 1 %
- Constant Value Column Filter< 1 %
- Nominal Value Row Splitter< 1 %
- Normalizer Apply (PMML)< 1 %
- Number to String (PMML)< 1 %
- String Replacer< 1 %
- Rule-based Row Splitter< 1 %
- Cache< 1 %
- Conditional Box Plot (legacy)< 1 %
- Crosstab< 1 %
- HiLite Table (legacy)< 1 %
- k-Medoids< 1 %
- Model Loop Start< 1 %
- DL4J Feedforward Predictor (Layer) (legacy)< 1 %
- Math Formula (Multi Column)< 1 %
- Java Snippet Row Splitter< 1 %
- Category to Class< 1 %
- Double Input (legacy)< 1 %
- Decision Tree View (JavaScript)< 1 %
- ROC Curve (JavaScript)< 1 %
- Scatter Plot (JavaScript)< 1 %
- Date&Time to String< 1 %
- Date Field Extractor (legacy)< 1 %
- Time Field Extractor (legacy)< 1 %
- Extract Time Window (legacy)< 1 %
- String to XML< 1 %
- Scorer< 1 %
Views
- This node has no views
Workflows
01_AnalyticsKNIME Hub- 02_Model_BuildingKNIME Hub
- 05_Gradient_Boosted_TreesKNIME Hub
- 05_Gradient_Boosted_TreesKNIME Hub
- 05_Gradient_Boosted_TreesKNIME Hub
- Show all 15 workflows
Links
- No links available
Developers
You want to see the source code for this node? Click the following button and we’ll use our super-powers to find it for you.