MAP Values: Federated Learning Algorithm Performance on Multimodal Benchmarks

Li Wan published a federated learning framework for multimodal data fusion on figshare in May 2026. The dataset likely contains algorithm performance metrics, specifically Mean Average Precision (MAP) values, from experiments on the TREC2017 Precision Medicine Track and CMU-MOSI sentiment benchmarks. The file is 5.5 KB in size.

Use Cases

• Benchmark federated learning algorithms based on reported MAP values
• Compare multimodal fusion methods based on performance on text-dominant and multimodal sentiment tasks
• Evaluate privacy-aware training frameworks based on the described decentralized approach
• Analyze the impact of tensor Tucker decomposition on modeling intermodal correlations

Strengths

• File size is 5.5 KB, indicating a compact and focused dataset
• Results are validated on two established benchmarks: TREC2017 Precision Medicine Track and CMU-MOSI
• Framework addresses specific challenges like privacy risks and parameter redundancy mentioned in the description

Limitations

• Row count is unknown, which may limit suitability assessment
• Column-level documentation is absent; field semantics must be inferred after download
• Description metadata is limited; actual data quality requires manual inspection after download

Provenance

Source

figshare

Collection Method

Likely experimental results from a proposed federated learning framework

Freshness

Last updated 2026-05-06 17:23:12; freshness should be verified