CS109: Introduction to Probability for Computer Scientists

Welcome to CS109

Instructor: Chris, professor in the Computer Science Department and School of Education.
Courses taught: CS109, CS106A, CS398, and more.
Background: Born in Kenya, moved to Malaysia, then to the US.
Research focus: AI for social good, education, medical work.
Experience: Taught 20,000 students during the pandemic, using AI to analyze code and provide feedback.

Website: cs109.stanford.edu.
Prerequisites:
- CS106B: Important for understanding coding concepts like recursion, hash tables, binary trees.
- CS103: Less critical; used lightly.
- Math 51/CME 100: Important for partial derivatives and multivariate integration.
Units:
- Undergraduates: 5 units.
- Graduates: Option for fewer units, but workload remains the same.
Components: Assignments, midterm, and final exams.
Attendance: Encouraged, but lectures are recorded.

Major components:
- Assignments: Practice and apply probability concepts.
- Exams: Midterm and final, focus on learning rather than judging.
- Sections: Small group meetings with TAs.
Participation: Incentivized through slight grade adjustments for attending lectures.

Modern AI: Combines probability and programming.
Historical context:
- 1950s: Initial optimism about AI capabilities.
- AI winter: Funding and interest declined due to overestimation of capabilities.
Recent advancements:
- Chess, self-driving cars, Jeopardy, protein folding.
- Large language models and AI systems.
- Applications in daily life: Navigation, translation, image generation.

Neural Networks:
- Mimic neurons through inputs, weights, and outputs.
- Learning by examples: Adjust weights based on training data.
Learning by Examples:
- Use examples to adjust weights and improve accuracy.
- Probability and programming underpin these advancements.

Critical for understanding AI, algorithms, and real-world phenomena.
One-Shot Learning: Highlighting current limitations (e.g., recognizing symbols with minimal examples).
Applications in decision-making and understanding uncertainties.

Ongoing challenges: Fairness in AI, ethical considerations, and applications in diverse fields.
CS109 Objectives: Teach probability to enable solutions to real-world problems.

Counting Theory:
- Experiment: An action with possible outcomes.
- Step Rule of Counting: Break down into steps and multiply outcomes.
- Sum Rule of Counting: Add outcomes from mutually exclusive sets.
- Inclusion-Exclusion Principle: Adjust for overlapping outcomes in sets.