Understanding the computational difficulty of a binary-weight perceptron and the advantage of input sparseness

Mathematics

• Perceptron 93%
• Binary 52%
• Decimation 47%
• Cross-correlation 12%
• Condensation 12%
• Neural Networks 9%
• Sparse Coding 9%
• Learning 9%
• Hardware Implementation 7%
• Pruning 6%
• Accessibility 6%
• Brain 5%
• Neuron 5%
• Assign 4%
• Efficient Algorithms 4%
• Output 3%
• Subspace 3%
• Modeling 3%
• Model 1%

Physics & Astronomy

• self organizing systems 100%
• low weight 16%
• learning 12%
• cross correlation 11%
• condensation 10%
• transition points 6%
• fixing 5%
• neurons 5%
• hardware 5%
• brain 5%
• coding 5%
• output 3%

Engineering & Materials Science

• Neural networks 42%
• Condensation 23%
• Preferred numbers 20%
• Neurons 10%
• Brain 9%
• Computer hardware 6%