How debarghabiswas/ml_practice works
Main Components
• Central “Machine Learning Course” or “Learning Modules” container – representing the overarching curriculum.
• Individual learning modules, represented as major components:
• Data Preprocessing (shows initial data manipulation and cleaning)
• Association Rule Learning (e.g., Apriori)
• Classification (with subcomponents for Decision Tree, K-NN, Naive Bayes, Support Vector Machines, Kernel SVM, Logistic Regression, Random Forest, etc.)
• Clustering (with Hierarchical and K-Means)
• Regression (with submodules for Simple Linear, Multiple Linear, Polynomial, Decision Tree, Random Forest, Support Vector Regression, and a branch for Regression Model Selection)
• Natural Language Processing (focused on textual data processing)
• Reinforcement Learning (with subcomponents for Thompson Sampling and UCB)
Data Flow and Interactions
• Data Preprocessing Component: Capture this as a pre-step that prepares raw data for further analysis. Indicate that outputs (cleaned/processed data) could feed into multiple ML modules.
Architectural Patterns and Design Principles
• Modular Design: Emphasize that each learning module is self-contained, demonstrating the principle of separation of concerns. Each module focuses on a specific algorithm or family of algorithms.
Technologies and Tools
• Jupyter Notebooks and Python Scripts: Represent these as execution environments/containers for each module.
• Data Files (CSV/TSV): Illustrate these as persistent data sources.
• Underlying Python libraries (implied by ML tasks, though not explicitly listed, one might expect libraries such as scikit-learn, pandas, numpy, etc.) can be mentioned if needed for further clarity.