Coding-For-MBA

Featured Lessons

This page highlights key lessons that demonstrate advanced concepts and production-ready patterns.

MLOps and Production Workflows

Day 50 – MLOps

File: Day_50_MLOps/solutions.py

Exposes reusable helpers for training, saving, loading, and predicting with a Logistic Regression Iris classifier.

Quick start:

python Day_50_MLOps/solutions.py

Programmatic usage:

from Day_50_MLOps.solutions import (
    load_model,
    predict_sample,
    save_model,
    train_iris_model,
)

model, accuracy, X_test, y_test, target_names = train_iris_model()
model_path = save_model(model, "iris_model.joblib")
reloaded = load_model(model_path)
prediction, label = predict_sample(reloaded, X_test[0], target_names)

Day 65 – MLOps Pipelines and CI/CD

File: Day_65_MLOps_Pipelines_and_CI/solutions.py

Implements CI/CD automation patterns for ML systems.

Day 66 – Model Deployment and Serving

File: Day_66_Model_Deployment_and_Serving/solutions.py

Demonstrates REST and gRPC serving patterns for production model deployment.

Day 67 – Model Monitoring and Reliability

File: Day_67_Model_Monitoring_and_Reliability/solutions.py

Covers monitoring, drift detection, and reliability engineering for ML systems.

Deep Learning and Modern NLP

Day 58 – Transformers and Attention

File: Day_58_Transformers_and_Attention/solutions.py

Builds encoder–decoder stacks, deterministic transformer text classifiers, Hugging Face fine-tuning playbooks, and attention visualisations for rapid experimentation.

Day 59 – Generative Models

File: Day_59_Generative_Models/solutions.py

Contrasts autoencoders, VAEs, GAN dynamics, and diffusion denoisers with synthetic training loops that log reconstruction improvements.

Day 64 – Modern NLP Pipelines

File: Day_64_Modern_NLP_Pipelines/solutions.py

Implements end-to-end NLP pipelines with state-of-the-art techniques.

Advanced ML Techniques

Day 60 – Graph and Geometric Learning

File: Day_60_Graph_and_Geometric_Learning/solutions.py

Implements GraphSAGE and graph attention message passing for toy node-classification graphs with interpretable attention matrices.

Day 61 – Reinforcement and Offline Learning

File: Day_61_Reinforcement_and_Offline_Learning/solutions.py

Simulates policy-gradient bandits, tabular value iteration, contextual bandits, and offline evaluation with deterministic reward thresholds for regression testing.

Day 62 – Model Interpretability and Fairness

File: Day_62_Model_Interpretability_and_Fairness/solutions.py

Covers SHAP, LIME, counterfactual explanations, and fairness metrics.

Day 63 – Causal Inference and Uplift

File: Day_63_Causal_Inference_and_Uplift/solutions.py

Demonstrates causal inference techniques and uplift modeling for business decisions.

Supervised and Unsupervised Learning

Day 51 – Regularised Models

File: Day_51_Regularized_Models/solutions.py

Compares ridge, lasso, and elastic net pipelines while introducing Poisson regression as a generalised linear model.

Day 52 – Ensemble Methods

File: Day_52_Ensemble_Methods/solutions.py

Trains random forests, gradient boosting, and stacking ensembles with calibration utilities for trustworthy probabilities.

Day 53 – Model Tuning & Feature Selection

File: Day_53_Model_Tuning_and_Feature_Selection/solutions.py

Demonstrates grid search, Bayesian optimisation (via skopt), permutation importance, and recursive feature elimination on a reproducible dataset.

Day 54 – Probabilistic Modeling

File: Day_54_Probabilistic_Modeling/solutions.py

Provides Gaussian mixtures, expectation-maximisation, Bayesian classifiers, and hidden Markov model log-likelihood utilities for reasoning under uncertainty.

Day 55 – Advanced Unsupervised Learning

File: Day_55_Advanced_Unsupervised_Learning/solutions.py

Explores DBSCAN, agglomerative clustering, t-SNE embeddings, autoencoders, and anomaly detection baselines.

Time Series and Recommenders

Day 56 – Time Series & Forecasting

File: Day_56_Time_Series_and_Forecasting/solutions.py

Fits ARIMA/SARIMAX, Holt-Winters, and Prophet-style models with rolling-origin evaluation helpers.

Day 57 – Recommender Systems

File: Day_57_Recommender_Systems/solutions.py

Implements collaborative filtering, matrix factorisation, and ranking metrics for implicit-feedback aware recommenders.

Data Engineering and Analytics

Day 31 – Relational Databases

File: Day_31_Databases/databases.py

Builds and queries a SQLite database, mirroring production-ready analysis workflows.

Day 32 – Other Databases

File: Day_32_Other_Databases/other_databases.py

Demonstrates dependency-injected connection patterns for SQL and MongoDB clients so that data access logic remains testable.

Day 36 – Capstone Case Study

File: Day_36_Case_Study/case_study.py and solutions.py

Ties together the full analytics workflow with a real dataset, cleaning, analysis, and business recommendations.

This site is open source. Improve this page.