Back To Course Home
Vector Databases: From Embeddings to Applications
0% completed
Before Getting Started
Introduction to the Course
Getting Started with Vector Databases and Embeddings
Introduction to Vector Databases and EmbeddingsMethods for Measuring Similarity between EmbeddingsEmbedding Models for Different Data TypesGenerating Text Data Embeddings with BERTGenerating Image, Video, and Audio EmbeddingsMultimodal EmbeddingsEmbeddings vs. Fine-TuningQuiz: Getting Started with Vector Databases and Embeddings
Working with Vector Databases
Overview of Open-Source Vector Databases
Practice Using ChromaDB for Multimodal Embeddings
Performance Optimization Techniques
Applications of Vector Databases
Quiz: Working with Vector Databases
Developing a Music Recommendation System
Introduction to the Music Recommendation SystemGenerating Embeddings for Audio Files and Metadata of SongsStoring Audio and Metadata Embeddings to Vector DatabaseQuerying the Database to Generate Recommendations
Wrapping Up
Conclusion

Performance Optimization Techniques

Learn about HNSW, a popular indexing method used in vector databases for efficient search.

We'll cover the following
  • The challenge of searching through large datasets
  • Indexing for faster searches
    • Hierarchical navigable small world (HNSW) graph
      • Navigable small world graph (NSW)
        • Building NSW graphs
        • Performing ANN search with NSW graphs
      • Skip lists
        • Search operation in a skip list
        • Insert operation in a skip list
        • Delete operation in a skip list
        • Benefits of skip lists
    • Integrating multi-level navigation from skip lists into NSW
  • Challenges of high-dimensional data
  • Data compression and dimensionality reduction
    • Product quantization (PQ)
    • Principal component analysis (PCA)

The challenge of searching through large datasets

Get hands-on with 1400+ tech skills courses.