MLE Masterclass

Month 1: Advanced Data Structures & Algorithms

• Week 1-2: Arrays, Strings, Two Pointers
  • Sliding window, prefix sums, cyclic sort.
  • ML Tie-in: Optimizing feature extraction (e.g., rolling averages).

• Post-Class Work:
  • LeetCode Medium/Hard (e.g., substring problems)

• Week 3-4: Trees, Graphs, Recursion
  • DFS/BFS, trie-based autocomplete, heap-based scheduling.
  • ML Tie-in: Decision tree pruning, graph embeddings.

Month 2: System Design (ML-Focused)

• Week 1-2: High-Level System Design
  • Load balancing, caching (Redis), CAP theorem.
  • Case Study: Design TinyURL → ML version: Design a feature store.
• Week 3-4: Distributed Systems for ML
  • Kafka for event-driven pipelines, sharding training data.
  • Project: Fault-tolerant model training on spot instances.

Month 3: Behavioral & Mock Interviews

• Week 1-2: Resume, LinkedIN & Mock Interviews
  • Resume and Linkedin update
  • 1:1 mock interviews 
  • Group whiteboarding sessions
  • Revise weak areas
• Week 3-4: Final Mock Interviews & Strategy
  • Full-loop mock interviews (coding + system design + behavioral)
  • Negotiation & offer discussion strategies
  • Final revision & relaxation before interviews

Month 4: Foundations of Machine Learning

• Week 1-2: Essential Mathematics for ML
  • Linear algebra, probability, and statistics
• Week 3-4: Supervised Learning
  • Regression and classification algorithms (e.g., Linear Regression, Logistic Regression, Decision Trees)
• Project: Data Analysis Pipeline
  • Develop a data analysis pipeline using Python, perform EDA, and visualize insights.

Month 5: Advanced Machine Learning Concepts

• Week 1-2: Unsupervised Learning
  • Clustering algorithms (e.g., K-means, Hierarchical Clustering)
  • Dimensionality reduction techniques (PCA, t-SNE)
• Week 3-4: Neural Networks and Deep Learning
  • Basics of neural networks
  • Introduction to TensorFlow/Keras
• Project: Predictive Model Development
  • Develop and train both a regression and a classification model. Evaluate and document the results.

Month 6: Data Handling and Preprocessing

• Week 1-2: Data Collection and Cleaning
  • Handling missing data
  • Data normalization and standardization
• Week 3-4: Exploratory Data Analysis (EDA)
  • Visualization techniques (Matplotlib, Seaborn)
  • Feature engineering
• Project: Comprehensive Data Cleaning and EDA
  • Perform thorough data cleaning, preprocessing, and EDA on a new dataset.

Month 7: Machine Learning Engineering in Practice

• Week 1-2: Model Evaluation and Validation
  • Cross-validation techniques
  • Metrics for classification and regression
• Week 3-4: Model Deployment and MLOps
  • Model serving (Flask, Docker)
  • Continuous Integration/Continuous Deployment (CI/CD) for ML
  • Monitoring and maintaining ML models in production
• Project: End-to-End ML Pipeline
  • Create a pipeline including data collection, model training, deployment, and monitoring.

Month 8: Capstone Project

• Week 1-2: NLP & LLMs
  • Fine-tuning Llama 3, RAG with Pinecone.
  • Horovod for distributed training, model parallelism.
  • Project: Train ResNet on 1TB of images (AWS SageMaker).
• Week 3-4: Recommender Systems
  • Collaborative filtering (ALS), two-tower models.
  • Capstone:
    • Amazon Recommender System (scraping → deployment → A/B testing).
    • Deliverables: Design doc, trained model, monitoring dashboard.