← 📊 Evaluation & ML Foundations
Core
Convex vs Non-Convex Optimization
A convex loss has one global minimum, so gradient descent from anywhere reaches the best solution: this is why logistic regression, linear regression, and linear SVMs are reliable to train. Deep nets are non-convex, with many local minima and far more saddle points, yet SGD still finds good solutions. Understanding why changes how you set initialization, learning rate, and restarts.
a free account unlocks the core curriculum tier · no card
RELATED CONCEPTS
PRACTICE THIS IN REAL QUESTIONS
ML Infrastructure & GPUsExplain memory coalescing and shared-memory bank conflicts. How would you fix a kernel that has both?→Machine Learning & Data ScienceExplain gradient descent to me like I'm a smart engineer who's never done ML. What can go wrong?→LLM & GenAI FundamentalsYour prompt plus retrieved documents exceed the context window. What are your options and tradeoffs?→ML Infrastructure & GPUsWhat do ZeRO and FSDP actually shard, and how much memory does each stage save? Where does gradient checkpointing fit?→ML Infrastructure & GPUsWe need to train a 100B-parameter model that won't fit in memory. Design the data and model parallelism.→SQL & Data EngineeringA customer's query on a 2-billion-row table takes 40 minutes. Walk me through your diagnosis.→