| layout | website | subdomain |
|---|---|---|
subsite-galaxy |
ml |
{:.no_toc}
{:.rna-intro-right}
The Galaxy Machine Learning workbench is a comprehensive set of analysis tools, consolidated workflows and training material. The workbench is based on the Galaxy framework{:target="_blank"}, which guarantees simple access, easy extension, flexible adaption to personal and security needs, and sophisticated analyses independent of command-line knowledge.
The current implementation provides you with the Swiss Army knife scikit-learn{:target="_blank"}, Keras{:target="_blank"} a deep learning library based on TensorFlow{:target="_blank"} and various other tools to manipulate, convert and plot your data.
The workbench is currently developed by the Goecks Lab and the European Galaxy project{:target="_blank"}. The German Network for Bioinformatics Infrastructure (de.NBI){:target="_blank"}, running the German ELIXIR Node{:target="_blank"} is providing the underlying CPU and GPU cluster.
This project is a community effort, please jump in, ask questions, contribute new tools, workflows or trainings!
{:.no_toc}
- TOC {:toc}
Are you new to Galaxy, or returning after a long time, and looking for help to get started? Take [a guided tour]({{ page.website }}/tours/core.galaxy_ui){:target="_blank"} through Galaxy's user interface.
We are passionate about training. So we are working in close collaboration with the Galaxy Training Network (GTN){:target="_blank"} to develop training materials of data analyses based on Galaxy {% cite batut2017community %}. These materials hosted on the GTN GitHub repository are available online at https://training.galaxyproject.org{:target="_blank"}.
Want to learn more about Machine Learning? Take one of our guided tours or check out the following hands-on tutorials. We developed several tutorials together with the GTN community.
In this section we list the most important tools that have been integrated in the Machine Learning workbench. There are many more tools available so please have a more detailed look into the tool panel. To ease readability, we divided them into categories.
Identifying which category an object belongs to.
| Tool | Description | Reference |
|---|---|---|
| {% include tool.html id="SVM Classifier" %} | SVM classifier | Pedregosa et al. 2011{:target="_blank"} |
| {% include tool.html id="NN Classifier" %} | Nearest Neighbors Classification | Pedregosa et al. 2011{:target="_blank"} |
| {% include tool.html id="Ensemble classification" %} | Ensemble methods for classification and regression | Pedregosa et al. 2011{:target="_blank"} |
| {% include tool.html id="Discriminant Classifier" %} | Linear and Quadratic Discriminant Analysis are two classic classifiers with a linear and a quadratic decision surface respectively | Pedregosa et al. 2011{:target="_blank"} |
| {% include tool.html id="Generalized linear" %} | Generalized linear models for classification and regression | Pedregosa et al. 2011{:target="_blank"} |
| {% include tool.html id="Train/Test evaluation" %} | Train, Test and Evaluation fit a model using part of dataset and evaluate using the rest | Pedregosa et al. 2011{:target="_blank"} |
| {% include tool.html id="Model Prediction" %} | Model Prediction predicts on new data using a preffited model | Chollet et al. 2011{:target="_blank"} |
| {% include tool.html id="Sample Generator" %} | Generate random samples with controlled size and complexity | Pedregosa et al. 2011{:target="_blank"} |
| {% include tool.html id="Estimator Attributes" %} | Estimator attributes get important attributes from an estimator or scikit object | Pedregosa et al. 2011{:target="_blank"} |
| {% include tool.html id="Stacking Ensemble Models" %} | Stacking Ensembles builds stacking, voting ensemble models with numerous base options | Pedregosa et al. 2011{:target="_blank"} |
| {% include tool.html id="Search CV" %} | Hyperparameter Search performs hyperparameter optimization using various SearchCVs | Pedregosa et al. 2011{:target="_blank"} |
| {% include tool.html id="Build Pipeline" %} | Pipeline Builder an all-in-one platform to build pipeline, single estimator, preprocessor and custom wrappers | Pedregosa et al. 2011{:target="_blank"} |
| {% include tool.html id="Feature selection" %} | Feature Selection module, including univariate filter selection methods and recursive feature elimination algorithm | Pedregosa et al. 2011{:target="_blank"} |
| {% include tool.html id="Fitted model evaluation" %} | Evaluate a Fitted Model using a new batch of labeled data | Pedregosa et al. 2011{:target="_blank"} |
| {% include tool.html id="Model fitting" %} | Fit a Pipeline, Ensemble or other models using a labeled dataset | Pedregosa et al. 2011{:target="_blank"} |
| {% include tool.html id="Train/Test splitting" %} | Split Dataset into training and test subsets | Pedregosa et al. 2011{:target="_blank"} |
| {: .table.table-striped .tooltable} |
Predicting a continuous-valued attribute associated with an object.
| Tool | Description | Reference |
|---|---|---|
| {% include tool.html id="Ensemble regression" %} | Ensemble methods for classification and regression | Pedregosa et al. 2011{:target="_blank"} |
| {% include tool.html id="Generalized linear" %} | Generalized linear models for classification and regression | Pedregosa et al. 2011{:target="_blank"} |
| {: .table.table-striped .tooltable} |
Automatic grouping of similar objects into sets.
| Tool | Description | Reference |
|---|---|---|
| {% include tool.html id="Numeric clustering" %} | Different numerical clustering algorithms | Pedregosa et al. 2011{:target="_blank"} |
| {: .table.table-striped .tooltable} |
Comparing, validating and choosing parameters and models.
| Tool | Description | Reference |
|---|---|---|
| {% include tool.html id="Model validation" %} | Model Validation includes cross_validate, cross_val_predict, learning_curve, and more | Pedregosa et al. 2011{:target="_blank"} |
| {% include tool.html id="CLF Metrics" %} | Calculate metrics for classification performance | Pedregosa et al. 2011{:target="_blank"} |
| {% include tool.html id="Pairwise Metrics" %} | Evaluate pairwise distances or compute affinity or kernel for sets of samples | Pedregosa et al. 2011{:target="_blank"} |
| {% include tool.html id="Regression metrics" %} | Calculate metrics for regression performance | Pedregosa et al. 2011{:target="_blank"} |
| {: .table.table-striped .tooltable} |
Feature extraction and normalization.
| Tool | Description | Reference |
|---|---|---|
| {% include tool.html id="Data preprocessing" %} | Preprocess raw feature vectors into standardized datasets | Pedregosa et al. 2011{:target="_blank"} |
| {: .table.table-striped .tooltable} |
Feature extraction and normalization.
| Tool | Description | Reference |
|---|---|---|
| {% include tool.html id="Batch Models" %} | Build Deep learning Batch Training Models with online data generator for Genomic/Protein sequences and images | Chollet et al. 2011{:target="_blank"} |
| {% include tool.html id="Model Builder" %} | Create deep learning model with an optimizer, loss function and fit parameters | Chollet et al. 2011{:target="_blank"} |
| {% include tool.html id="Model Config" %} | Create a deep learning model architecture using Keras | Chollet et al. 2011{:target="_blank"} |
| {% include tool.html id="Train and evaluation" %} | Deep learning training and evaluation conduct deep training and evaluation either implicitly or explicitly | Chollet et al. 2011{:target="_blank"} |
| {: .table.table-striped .tooltable} |
General data manipulation tools
| Tool | Description | Reference |
|---|---|---|
| {% include tool.html id="Table compute" %} | The power of the pandas data library for manipulating and computing expressions upon tabular data and matrices. | |
| {% include tool.html id="datamash operations" %} | datamash operations on tabular data | |
| {% include tool.html id="datamash transpose" %} | Transpose rows/columns in a tabular file | |
| {% include tool.html id="Visualization" %} | Machine Learning Visualization Extension includes several types of plotting for machine learning | Chollet et al. 2011{:target="_blank"} |
| {: .table.table-striped .tooltable} |
You have done the heavy lifting and now want to use your coding skills inside Jupyter or RStudio? Work on data
| Tool | Description | Reference |
|---|---|---|
| Jupyter{:target="_blank"} | Jupyter lab | |
| RStudio{:target="_blank"} | RStudio | |
| {: .table.table-striped .tooltable} |