vanishing gradient problem
machine learning model training problem of greatly diverging gradient magnitudes between earlier and later layers encountered when training neural networks with backpropagation
Press Enter · cited answer in seconds
0 sources
vanishing gradient problem
Summary
vanishing gradient problem ranks in the top 2% of general entities by monthly Wikipedia readership (267 views/month).[1]
Key Facts
- vanishing gradient problem is credited with the discovery of Sepp Hochreiter[2].
- vanishing gradient problem's subclass of is recorded as problem[3].
- vanishing gradient problem's opposite of is recorded as exploding gradient problem[4].
- vanishing gradient problem's time of discovery or invention is recorded as +1991-00-00T00:00:00Z[5].
- vanishing gradient problem's Freebase ID is recorded as /m/011lfk_8[6].
- vanishing gradient problem's facet of is recorded as machine learning[7].
- vanishing gradient problem's Microsoft Academic ID is recorded as 2778149865[8].
- vanishing gradient problem's Golden ID is recorded as Vanishing_gradient_problem-ERA89Y[9].
- vanishing gradient problem's Encyclopedia of China is recorded as 413695[10].
Body
Works and Contributions
vanishing gradient problem is credited with the discovery of Sepp Hochreiter[2].
Why It Matters
vanishing gradient problem ranks in the top 2% of general entities by monthly Wikipedia readership (267 views/month).[1] It has Wikipedia articles in 11 language editions, a strong signal of global cultural recognition.[11]