From Zero to Mastery: A Practical Complete Curriculum for Machine Learning and Generative AI

Core Information

• Course Name: DSAI Module3—10-Lesson Practical Course on Machine Learning and Generative AI
• Original Author: flexfengfeng
• Source Platform: GitHub (Link)
• Target Audience: Business analysts, product managers, technical leaders, and professionals who want to expand their Python and AI skills
• Core Value: Adopts a narrative learning path (following Sarah Chen's career experience), with an 80% practice + 20% theory ratio, covering the full skill set from classic ML to Transformer and RAG, with a low entry barrier (only basic Python/SQL required).

Background

With the popularization of artificial intelligence, non-technical professionals (such as business analysts) understand the potential of AI but lack a systematic learning path. Traditional courses are too academic, filled with complex formulas and abstract concepts.

Design Philosophy

Solve pain points with the 'narrative-driven learning method': Combine technical knowledge with real business scenarios through the career experience of the fictional character Sarah Chen (Customer Experience Analyst at NorthStar Retail), lowering the learning threshold and allowing learners to understand the value of technical applications in real business.

Structure Ratio

Follows the golden ratio of 80% practice and 20% theory, ensuring hands-on coding in each lesson.

Learning Stages

Adopts a three-stage model: Pre-class self-study (about 75 minutes) → In-class practice (about 3 hours) → After-class assignments (completed independently).

Content Progression

Starts from basic ML concepts and gradually deepens:

1. Classic ML (probability statistics, supervised learning)
2. Unsupervised learning and time series analysis
3. Deep learning (neural networks, CV)
4. Natural language processing and generative AI (Transformer, RAG) Each lesson revolves around Sarah's business challenges, giving technology specific application scenarios.

Classic ML Module (Lessons 1-4)

• Build an ML cognitive framework and understand applicable scenarios of technology
• Supplement basic probability statistics (confidence intervals, A/B testing)
• Dive into supervised learning (logistic regression, gradient boosting trees, including feature engineering and model selection)

Expansion Module (Lessons5-6)

• Unsupervised learning (clustering, PCA) to discover hidden patterns in data
• Time series analysis for sales forecasting and inventory management

Deep Learning Module (Lessons7-8)

• Build MLP and CNN with PyTorch
• Focus on transfer learning (practical for small dataset scenarios)

Generative AI Module (Lessons9-10)

• NLP basics: word embedding, semantic search
• Transformer and RAG technologies to build intelligent customer service and document Q&A systems

Technical Tools

• Environment management: Miniconda
• Development environment: VS Code + Jupyter extension
• Deep learning framework: PyTorch

Cross-Platform Support

• macOS (including Apple Silicon)
• Windows: WSL2 is recommended (better PyTorch stability/performance)

GPU Alternative

For modules requiring GPU acceleration (such as CV, Transformer), Google Colab is provided as an alternative, making it convenient for learners without local GPUs to complete experiments.

End-to-End Projects

Learners need to choose real business problems (can use NorthStar cases or their own work scenarios), apply skills from L01-L10 to build a complete solution, and submit runnable code and documents.

Team Learning Support

Provides a hackathon guide (HACKATHON_GUIDE.md), including theme suggestions, review criteria, and activity arrangements, to meet the team collaboration needs of enterprise training.

Evaluation Flexibility

Supports individual in-depth learning and team collaboration evaluation, adapting to different learning scenarios.

Unique Methodology

• Narrative learning: Provide cognitive anchors through the character Sarah Chen, attach technology to business stories, and improve memory and transfer application capabilities
• Experiential learning: Follow the 'try → reflect → understand → apply' cycle to cultivate the ability to solve unfamiliar problems

AI Education Insights

• Trend: Shift from training AI researchers to training AI applicators
• Core: Good AI education requires clear business insights and continuous hands-on practice; you can get started without a deep mathematical background.