Machine Learning Algorithms & Pipelines

[KMouratidis]

1. Add more algorithms:

• [x] sklearn (46 so far)
• [ ] statsmodels
• [ ] tensorflow / pytorch
• [ ] NLP models (nltk, gensim, spacy, etc)

2. Implement common pipelines, e.g.:

• [ ] Imputation -> Standardization -> Feature extraction -> Training
• [ ] Voting Classifiers, ensembles & other aggregators
• [ ] Cross-validation, (hyper-)parameter tuning
• [ ] Dimensionality reduction

3. Graphical model designer

• [x] Basic layout
• [x] Basic conversion to sklearn.pipeline.Pipeline
• [x] Redefine edges, restyle arrows
• [x] Advanced functionality
• [x] Handle order better