Smoothed InfoNCE: Breaking the log N Curse without Overshooting.
Xu WangAli Al-BashabshehChao ZhaoChung ChanPublished in: ISIT (2022)
Keyphrases
- high dimensional
- lower bound
- high dimensional data
- dimensionality reduction
- worst case
- high dimensionality
- dimension reduction
- training data
- upper bound
- hidden markov models
- principal component analysis
- np hard
- artificial intelligence
- mobile devices
- artificial neural networks
- multi agent systems
- image sequences
- feature selection