Build a Random Forest Classifier using CellProfiler Image features (Master DO NOT UPLOAD)
This workflow is used to train a random forest model to classify images as focused or unfocused. This has been designed to work with image features extracted in CellProfiler.
Extracted features from training images are read into the workflow. This data is annotated with a class variable and partitioned between training and validation. Hyperparameter optimization is carried out and parameters which perform the best are selected and used to train a deployment model. Features extracted from the deployment images are read into the workflow and classified. Predictions are saved to an Excel file.
This workflow contains the following components:
(1) Training data annotation
(2) Training and validation
(3) Hyperparameter optimization
(4) Deployment
URL: Decision Tree Node: Algorithm Settings https://youtu.be/CSwM92yTrJw
Nodes
- Column Rename3 ×
- DB Query Reader (Labs)3 ×
- Random Forest Learner3 ×
- Random Forest Predictor3 ×
- SQLite Connector (Labs)3 ×
- Show all 15 nodes
Extensions
- FeatureKNIME Base nodes
- FeatureKNIME Database
- FeatureKNIME Ensemble Learning Wrappers
- FeatureKNIME Excel Support
- FeatureKNIME Javasnippet
- Show all 6 modules
Links
- No links available