4 .Deep Learning and Embedding
In the last years, deep learning has made signi cant progress
in speech recognition, computer vision, and natural language understanding.
Among them embedding, which is also called
representation learning, has been proven to be successful techniques
contributing to the success. [1]
[1] Jui-Ting Huang et al. “Embedding-based Retrieval in Facebook Search” knowledge discovery and data mining (2020): n. pag.
fi
5 .Deep Learning and Embedding
6 .Deep Learning and Embedding
7 .Deep Learning and Embedding
8 .Deep Learning and Embedding
[1] Jui-Ting Huang et al. “Embedding-based Retrieval in Facebook Search” knowledge discovery and data mining (2020): n. pag.
9 .Deep Learning and Embedding
10 .Deep Learning and Embedding
[1] Jui-Ting Huang et al. “Embedding-based Retrieval in Facebook Search” knowledge discovery and data mining (2020): n. pag.
11 .Fast Growth of Unstructured Data
12 .Unstructured Data + Embedding
13 .Scenarios
• Recommendations System
• Search and retrieval of images (CV related scenarios)
• Text to vector (NLP related scenarios)
• Finding audio/songs that are similar
• Protein Sequencing
• Reduce the dimensions of high-dimensional input data
• …
27 .Milvus——向量数据库（Vector Database）
A vector database indexes and stores vector embeddings for fast retrieval and similarity search
• Manage vector data - CRUD, Import/Export, Backup, Snapshot
• Scalability
• Availability
• Persistence
• Data freshness
• Easy to use - Open-box usage experience
热门城市
