CSCI-B 559 Efficient Artificial Intelligence
3 credits
- Prerequisite(s): None
- Delivery: On-Campus
Description
This course covers efficient artificial intelligence theories and techniques for designing advanced machine learning models. The goal is to optimize architecture, implementation, performance, power, memory, and time. Students learn efficient AI theories, patterns in AI, critical pattern determination, AI complexity, AI optimization, large deep models, real-time inference, and practical applications.
Topics
Introduction of efficient AI and deep learning
- Challenges in large AI models
- Challenges in big data learning
- Deep learning and machine learning
- Various deep learning architectures
- Algorithm and model implementation
- Computation efficiency
- Power efficiency
Efficient pattern extraction in AI
- Data types
- Patterns in the big data
- Data preprocessing
- Data cleaning
- Data integration and transformation
- Data reduction
- Define pattern efficiency in the learning process
- Engineering pattern efficiency for deep and machine learning models
Efficient pattern determination in AI
- Various pattern types and evaluation
- Pattern ranking without or with AI models
- Statistical methods
- Regularization methods
- Sparse learning methods
- Pattern determination for efficient AI learning
- Quantization for AI models
Efficient AI model design
- Deep learning models
- Learning theories
- Neural network and pruning
- Iterative pruning
- Pruning criteria and principles
- Neural network sparsification
- Model granularity
- Model complexity
- Model efficiency
- Algorithm-implementation
- AI model deployment and evaluation
AI Model execution
- Large deep learning models
- Distributed training principles
- AI model inference
- Time efficiency
- Power efficiency
- Large-scale models
- Various deep learning applications
Learning Outcomes
- Analyze and compare major methods for efficient pattern extraction and determination.
- Assess the performance of various efficient pattern extraction and determination techniques.
- Design and simulate the AI model, focusing on minimizing complexity and maximizing efficiency.
- Understand and examine AI models that optimize both time and power efficiency.
- Design and implement the efficient neural network model, with a course project.
Policies and Procedures
Please be aware of the following linked policies and procedures. Note that in individual courses instructors will have stipulations specific to their course.