| DL & Systems | Implementations | Core Concepts |
|---|---|---|
| Backpropagation | 0/1 | Chain Rule, computational graph, Jacobian matrix calculation. |
| CNNs (Convolutions) | 0/3 | Kernel operation, stride, padding, pooling layers. |
| Transformers & Attention | 0/3 | Scaled Dot-Product, Positional Encoding, Multi-Head structure. |
| Regularization & Optimization | 0/3 | Dropout, L1/L2 penalties, Adam/SGD weight update formulas. |
| Model Optimization (Inference) | 0/3 | Quantization (FP16/INT8), Pruning, Latency vs. Throughput. |
| ML Systems & MLOps | Implementations | Core Concepts |
|---|---|---|
| System Design | 0/3 | Designing end-to-end ML pipelines for large-scale systems. |
| Inference & Deployment | 0/3 | Model serving, A/B testing, API endpoints, microservices. |
| Data Pipelines | 0/3 | Data ingestion, validation, feature stores, ETL workflows. |
| ML Algorithm / Concept | Implementations | Core Concepts |
|---|---|---|
| Model Evaluation | 0/3 | Precision, Recall, F1-Score, and AUC. |
| Linear & Logistic Regression | 0/3 | Gradient Descent, loss functions, classification/regression. |
| K-Means & Clustering | 0/3 | Centroid initialization and distance metrics. |
| PCA & Dimensionality Reduction | 0/3 | Eigenvectors/Eigenvalues and variance. |
| Category | Implementations | Core Concepts |
|---|---|---|
| LLM Specifics | RAG (Retrieval-Augmented Generation), Hallucination Prevention | Vector Databases, Indexing, Embeddings, Prompt Engineering, Citation Mechanisms, Fine-tuning vs. In-Context Learning |
| System Design | API Endpoints (REST/gRPC), Load Balancers, Caching Layers | Designing End-to-End LLM Pipelines, Scalability, High Availability, Fault Tolerance, Latency vs. Throughput Trade-offs |
| MLOps & Production | Model Monitoring, Drift Detection, A/B Testing, Shadow Deployment | Data and Concept Drift, Feature Stores, Model Versioning, CI/CD for ML (MLOps pipelines) |
| Category | Implementations | Core Concepts |
|---|---|---|
| Applied Problem-Solving | Project Discussions, Technical Decision-Making, Trade-off Analysis | The STAR method (Situation, Task, Action, Result), Handling Failure, Technical Communication with Non-Experts |
| Foundational ML | Feature Engineering, Data Preprocessing | Bias-Variance Trade-off, Overfitting, Underfitting, Model Selection, Unsupervised Learning (e.g., Clustering, Dimensionality Reduction) |