This node performs supervised training of a feedforward deep learning model for classification. Thereby, the learning procedure
can be adjusted using several training methods and parameters, which can be customized in the node dialog. Additionally,
the node supplies further methods for regularization, gradient normalization and learning refinements. The learner
node automatically adds an output layer to the network configuration, which can be also configured in the node
dialog. For classification, the output layer will always use 'softmax' as the activation function and the number
of outputs will be automatically set to match the number of unique labels used for training. The output of the
node is a trained deep learning model which can be used to predict the labels of unseen data instances.
The KNIME Deeplearning4J Integration has been marked as legacy with KNIME Analytics Platform 5.0 and will be deprecated in a future version. If you are using this extension in a production workflow, consider switching to one of the other deep learning integrations available in KNIME Analytics Platform.
Options
Learning Parameters
Global Parameters
Data Parameters
Column Selection
Output Layer Parameter
Popular Predecessors
- Missing Value5 %
Number To String5 %DL4J Model Reader4 %- Shuffle4 %
File Reader3 %CSV Reader2 %- Data Generator2 %
- Table To DL4J Vector (Classification) (legacy)2 %
Column Appender2 %RBM Layer (legacy)2 %- DL4J Feedforward Predictor (Classification) (legacy)2 %
- Read Images2 %
- LRN Layer (legacy)2 %
- Normalizer (Apply)2 %
Excel Reader (XLS)2 %- Java Snippet1 %
String To Number1 %- Dense Layer (legacy)1 %
- Number to String1 %
- DL4J Feedforward Learner (Pretraining) (legacy)1 %
- Word Vector Apply (legacy)1 %
- Scatter Plot (JavaScript)1 %
- DL4J Model Reader (legacy)1 %
- AutoEncoder Layer (legacy)1 %
- DL4J Model Initializer (legacy)1 %
- DL4J Feedforward Predictor (Layer) (legacy)1 %
- PCA1 %
- SMOTE< 1 %
Row Splitter (deprecated)< 1 %- DL4J Feedforward Learner (Regression) (legacy)< 1 %
- Image Normalizer< 1 %
- Joiner< 1 %
- Split Collection Column< 1 %
Database Reader< 1 %Counting Loop Start< 1 %- Auto-Binner< 1 %
- Expand Bit Vector< 1 %
- Category to Number< 1 %
- Domain Calculator< 1 %
- Normalizer< 1 %
- Partitioning< 1 %
- RowID< 1 %
- Rule-based Row Filter< 1 %
- Color Manager< 1 %
- Concatenate< 1 %
- Edit Nominal Domain< 1 %
- Column Filter< 1 %
Row Filter (deprecated)< 1 %- Convolution Layer (legacy)< 1 %
- Pooling Layer (legacy)< 1 %
- DL4J Feedforward Learner (Classification) (legacy)< 1 %
- Table To DL4J Vector (Pretraining) (legacy)< 1 %
- Image to DataRow< 1 %
Table Reader< 1 %- Column Combiner< 1 %
- Extract Missing Value Cause< 1 %
- Nominal Value Row Filter< 1 %
CASE Switch Data (Start)< 1 %- ROC Curve (legacy)< 1 %
- Fingerprints< 1 %
- Reference Row Splitter< 1 %
- String Manipulation< 1 %
- Boosting Learner Loop Start< 1 %
- Convolution Kernel Creator< 1 %
- Fingerprint to String< 1 %
- MultiLayerPerceptron Predictor< 1 %
- Bootstrap Sampling< 1 %
- Correlation Filter< 1 %
- Equal Size Sampling< 1 %
Column Rename< 1 %- Sorter< 1 %
- SDF Reader< 1 %
- Math Formula< 1 %
Excel Reader (XLS)< 1 %- Global Thresholder< 1 %
- Fingerprint Similarity< 1 %
Table Row To Variable Loop Start< 1 %- Constant Value Column< 1 %
- Edit Numeric Domain< 1 %
- Missing Value Column Filter< 1 %
Number To String (PMML)< 1 %- Conditional Label Assigner< 1 %
DL Network Executor< 1 %- Gaussian Convolution< 1 %
- Feature Calculator (BETA)< 1 %
- RDKit Descriptor Calculation< 1 %
- Benchmark Start< 1 %
- Fingerprints Expander< 1 %
- Create Collection Column< 1 %
Table Row to Variable< 1 %- ARFF Reader< 1 %
- Table Creator< 1 %
- X-Partitioner< 1 %
Scorer< 1 %Cell Replacer< 1 %- One to Many< 1 %
- Double to Integer< 1 %
- GroupBy< 1 %
Joiner< 1 %- Low Variance Filter< 1 %
- Missing Value (Apply)< 1 %
- Row Sampling< 1 %
- Column Splitter< 1 %
- String Manipulation (Variable)< 1 %
- Table Transposer< 1 %
- Rule Engine< 1 %
Color Appender< 1 %- Java Edit Variable (simple)< 1 %
- Keygraph Keyword Extractor< 1 %
- Data Row to Image< 1 %
- Image Resizer< 1 %
- Slice Loop Start< 1 %
- Parameter Optimization Loop Start< 1 %
- Fingerprint to Set Bits List< 1 %
- Chemical Terms< 1 %
- PNGs to Binary Objects< 1 %
- Strings to Binary Objects< 1 %
Table Column to Variable< 1 %Model Reader< 1 %- Java Edit Variable< 1 %
- Feature Selection Filter< 1 %
- Feature Selection Loop End< 1 %
- Loop End (Column Append)< 1 %
PCA Apply< 1 %PCA< 1 %- Create Bit Vector< 1 %
- Column Auto Type Cast< 1 %
- String to Number< 1 %
- Column Resorter< 1 %
- Reference Row Filter< 1 %
- One to Many (PMML)< 1 %
- Normalizer (PMML)< 1 %
- Target Shuffling< 1 %
- Parallel Chunk Start< 1 %
- PMML Ensemble to Table< 1 %
- PMML to Cell< 1 %
Read Excel Sheet Names (XLS)< 1 %- Vocabulary Extractor (legacy)< 1 %
- TF< 1 %
- Chi-Square Keyword Extractor< 1 %
- Column Filter (legacy)< 1 %
- Integer Input (legacy)< 1 %
- String Input (legacy)< 1 %
- Table View (JavaScript)< 1 %
- JSON to Table< 1 %
- Image Generator< 1 %
- Labeling to String based Labeling< 1 %
- Dimension Swapper< 1 %
- Image Converter< 1 %
- Image Cropper< 1 %
- Inverter< 1 %
- Binary Image Operations< 1 %
- Connected Component Analysis< 1 %
- Labeling Converter< 1 %
- Image to Labeling< 1 %
- Salt and Pepper< 1 %
Python Script (1⇒1)< 1 %Column Selection QuickForm (legacy)< 1 %- SPARQL Query< 1 %
- Test FileStore Column< 1 %
- Create FileStore Column in LoopEnd< 1 %
- Time Generator (legacy)< 1 %
- RDKit Count-Based Fingerprint< 1 %
- CSV Reader< 1 %
- Shape Appender< 1 %
- Shape Manager< 1 %
- Size Manager< 1 %
Views
- Learning Status
- Shows information about the current learning run. Has an option for early stopping of training. If training is stopped before the last epoch the model will be saved in the current status.
Workflows
01_AnalyticsKNIME Hub- 01_Basic_Concepts_Of_Deeplearning4J_IntegrationKNIME Hub
- 01_Using_DeepLearning4J_to_classify_MNIST_DigitsKNIME Hub
- 01_Using_DeepLearning4J_to_classify_MNIST_DigitsKNIME Hub
- 02_Basic_Learner_View_TutorialKNIME Hub
- Show all 21 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.