t-SNE is a manifold learning technique, which learns low dimensional embeddings for high dimensional data. It is most often used for visualization purposes because it exploits the local relationships between datapoints and can subsequently capture nonlinear structures in the data. Unlike other dimension reduction techniques like PCA, a learned t-SNE model can't be applied to new data. The t-SNE algorithm can be roughly summarized as two steps:
- Create a probability distribution capturing the relationships between points in the high dimensional space
- Find a low dimensional space that resembles the probability dimension as well as possible
Disclaimer:
Depending on the size of the input table, the computation of t-SNE can be very expensive both in terms of runtime as well as memory. If you experience problems with memory, try to reduce the size of your data by e.g. using the Row Sampling node. If you have very high-dimensional data, it is also advisable to first reduce the number of dimensions to around 50 using e.g. a PCA.Options
Popular Predecessors
- Concatenate6 %
- Column Merger3 %
- Column Filter3 %
- Normalizer3 %
- Row Sampling3 %
- Table Transposer3 %
Row Splitter (deprecated)2 %- t-SNE (L. Jonsson)2 %
- Create Bit Vector2 %
- Fingerprints Expander2 %
Table Reader2 %- Normalizer (Apply)2 %
- Expand Bit Vector2 %
- Correlation Filter2 %
- Parameter Optimization Loop Start2 %
- Fuzzy c-Means1 %
- k-Means1 %
- PCA Apply1 %
- PCA1 %
- Nominal Value Row Filter1 %
Row Filter (deprecated)1 %- Missing Value1 %
- RowID1 %
- Color Manager1 %
- Topic Extractor (Parallel LDA)1 %
- Document Vector1 %
- RDKit Descriptor Calculation1 %
- CSV Reader1 %
- File Reader1 %
- Column Aggregator1 %
- Joiner1 %
- Column Splitter1 %
- k-Medoids1 %
- Excel Reader1 %
- RDKit Fingerprint1 %
- Loop End1 %
- String to Number1 %
- Distance Matrix Calculate1 %
- Split Collection Column1 %
- Partitioning1 %
- Table Row to Variable Loop Start1 %
- Denormalizer1 %
- Chi-Square Keyword Extractor1 %
- Low Variance Filter< 1 %
- Cache< 1 %
- BERT Embedder< 1 %
Table Row to Variable< 1 %CSV Reader< 1 %File Reader< 1 %Joiner< 1 %Merge Variables< 1 %Excel Reader (XLS)< 1 %- Weka Cluster Assigner (3.7)< 1 %
Column Filter Widget< 1 %- Reference Column Filter< 1 %
- Missing Value Column Filter< 1 %
- String Manipulation (Multi Column)< 1 %
- Rule Engine< 1 %
- Math Formula (Variable)< 1 %
- Scatter Plot (JavaScript)< 1 %
- Tag Cloud (JavaScript)< 1 %
- Normalize Plates (Z-Score)< 1 %
Loop End< 1 %- Hierarchical Clustering< 1 %
- DBSCAN< 1 %
- MDS< 1 %
- PCA Compute< 1 %
- Category to Number< 1 %
- Linear Correlation< 1 %
- Duplicate Row Filter< 1 %
- One to Many (PMML)< 1 %
- Number Rounder< 1 %
- Sorter< 1 %
- Data Explorer< 1 %
- Wilcoxon Signed-Rank< 1 %
- DB Query Reader< 1 %
- Counter Generation< 1 %
- Random Label Assigner (Data)< 1 %
- K Nearest Neighbor (Distance Function)< 1 %
- TensorFlow 2 Network Executor< 1 %
- 3D Scatter Plot (Plotly)< 1 %
Excel Reader (XLS)< 1 %- Document Vector Hashing< 1 %
- Column Filter (legacy)< 1 %
Python Script (1⇒1)< 1 %- Python Source (legacy)< 1 %
RDKit Fingerprint Reader< 1 %- Normalize Plates (POC)< 1 %
- Create Collection Column< 1 %
- Hierarchical Cluster Assigner< 1 %
- Number to String< 1 %
- HiLite Row Splitter< 1 %
- GroupBy< 1 %
- Data Generator< 1 %
- DB Reader< 1 %
- Mahalanobis Distance< 1 %
- Keras Network Executor< 1 %
- Scatter Plot (Plotly)< 1 %
- H2O to Table< 1 %
- Math Formula< 1 %
- Stop Word Filter< 1 %
Views
- This node has no views
Workflows
(3D) Fragment library design workflowKNIME Hub- 01_BERT_Sentiment_AnalysisKNIME Hub
- 02_PCA_t-SNEKNIME Hub
- 02_Techniques_for_Dimensionality_ReductionKNIME Hub
- 3.3_ChemicalSpaceKNIME Hub
- Show all 38 workflows
Links
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.