Workflow Element Store

  1. Handling Imbalanced Classes
  2. Handling Missing Data
  3. Handling Categorical Data
  4. Time-Based Features
  5. Textual Feature Extraction
  6. Feature Extraction from Images
  7. Domain-Specific Feature Engineering
  8. Auto-Preprocessing libraries
  9. Dealing with Outliers
  10. Handling Noisy Data
  11. Polynomial Features
  12. Data Scaling and Normalization
  13. Annotation
  14. Data Partitioning - Train, Validation, & Test
  15. Handling Time-Series Data
  16. Feature Selection
  17. AutoEDA libraries
  18. Augmentation
  19. Interaction Features
  20. Data Transformations
  21. Dimensionality Reduction
  22. Binning / Discretization
  1. Word Embeddings
  2. Natural Language Processing
  3. Transfer Learning
  4. Regular Monitoring and Logging
  5. Evaluation Metrics
  6. Data Augmentation
  7. GridSearchCV, RandomisedSearchCV, BayesianSearchCV
  8. External Validation
  9. Multiclass Classification Techniques
  10. Ensemble Techniques
  11. Hyperparameter Tuning
  12. Regularization Techniques
  13. Cross-Validation
  14. Learning Rate Scheduling
  15. Batch Normalization
  16. Batch Size Selection
  17. Network Analytics/ GeoSpatial Analytics
  18. Clustering
  19. Model Comparison
  20. Transfer Learning
  21. AutoML
  22. Regularization
  23. Recommendation Engine
  24. Binary Classification Techniques
  25. Model Interpretability
  26. Regression Analysis
  27. Cross-Validation
  28. Blackbox - Neural Network Models
  29. Forecasting Techniques
  30. Weight Initialization
  31. Association Rules
  32. Performance Visualization
  33. Reinforcement Learning
  34. Early Stopping
  1. Evidently.ai
  2. Apache Airflow
  3. Github
  4. Datawarehouse
  5. Data Preprocessing pipeline models
  6. Kafka Brokers
  7. model registry
  8. Databases
  9. code repository
  10. Github Actions
ML Workflow - Architecture
  • Element belongs to model
  • Element not belongs to model
Validate Export

Data Sources

Streaming Data

Batch Data

Cloud Storage

Labeled Data

Feature Engineering Pipeline

Experimentation

ML Model

Repository

CI/CD component

Continuous integration/Continuous delivery

Continuous deployment

Artifact Store

Feature Store System

Orchestration Component

Artifact Store

CI/CD Component

Scheduler

Workflow orchestration component

Automation ML Workflow Pipeline

Monitoring Component

Model serving component

(Prediction on new batch or streaming data)