Max - user overview - MetaOptimize Q+A

Max

476

karma

Registered user
member since	Jan 08 '13 at 02:28
last seen	May 04 '14 at 21:45

overview recent activity karma history favorites

15 Questions

votes

answers

581

views

Labeled image datasets illustrating verbs and adjectives?

May 01 '14 at 15:03 Max476●16●27●29

votes

answers

views

Are there free speech-recognition datasets?

May 04 '14 at 05:55 yesteve31●2 (asked by Max476●16●27●29 )

datasets dataset speech speech-recognition audio data

votes

answers

views

Bill Gates on Deep Learning at Microsoft

Feb 11 '14 at 22:03 eder216●2●5●11 (asked by Max476●16●27●29 )

deep-learning machine-learning

votes

answers

643

views

learning rate adjustments : short-term vs long-term benefits

Jan 28 '14 at 14:29 eder216●2●5●11 (asked by Max476●16●27●29 )

learning-rate deep-learning optimization neural-networks stochastic-gradient

votes

answers

838

views

Can Random Forests do much better than the 2.8% test error on MNIST?

Dec 06 '13 at 16:48 Max476●16●27●29

random-forests opencv r decision-trees machine-learning scikits-learn mnist

votes

answers

714

views

How well do Random Forests perform on the popular ML benchmarks?

Dec 05 '13 at 05:23 Max476●16●27●29

benchmark random-forests decision-trees ensemble mnist

votes

answers

views

Neural Network frameworks comparison: Torch7 vs Theano/Pylearn2 and others

Jun 26 '14 at 09:32 Nouiz61●2 (asked by Max476●16●27●29 )

torch pylearn2 libraries neural-networks theano pylearn software

votes

answers

views

Predictive Sparse Decomposition vs Sparse AutoEncoders?

May 29 '14 at 10:34 Christian Hudon1●2 (asked by Max476●16●27●29 )

autoencoder sparse-coding deep-learning neural-networks sparse unsupervised-learning

votes

answers

views

For what problems are rectified linear units better than tanh? Why are they better?

Nov 01 '13 at 18:08 Sander Dieleman1556●7●27●34 (asked by Max476●16●27●29 )

deep-learning rectified-linear-units neural-networks

votes

answers

views

Why isn't DropOut used in Unsupervised Learning?

Oct 31 '13 at 04:43 alfa86●3●7●12 (asked by Max476●16●27●29 )

deep-learning dropout neural-networks regularization unsupervised-learning

votes

answers

views

Better Mini-Batch Algorithms via Accelerated Gradient Descent

Oct 28 '13 at 09:54 Rakesh Chalasani264●1●2●10 (asked by Max476●16●27●29 )

paper-understanding gradient-descent stochastic-gradient optimization neural-networks learning algorithms sgd

votes

answers

views

Why would a "KL divergence" term enforce sparsity in autoencoders?

Oct 14 '13 at 17:35 Saul Berardo66●1●2●7 (asked by Max476●16●27●29 )

autoencoder deep-learning kl-divergence neural-networks sparse unsupervised-learning

votes

answers

views

Online Truncated Backpropagation Through Time efficiency

Sep 21 '13 at 14:14 Max476●16●27●29

backpropagation neural-networks rnn recurrent-neural-networks

votes

answers

views

How is divisive normalization a "poor man's whitening"?

May 15 '13 at 21:18 Alexandre Passos ♦25541●54●278●421 (asked by Max476●16●27●29 )

deep-learning computer-vision pca neural-networks

votes

answers

views

HMMs vs RNNs on real-world data

Feb 11 '13 at 06:38 Oscar Täckström2039●13●34●50 (asked by Max476●16●27●29 )

hmm neural-networks rnn recurrent-neural-networks

5 Answers

Anyone recommend a good, easy to use, hyperparameter-free adaptive learning rate for SGD?

Dropout nets: Why scaling weights after pretraining?

Single pass / Streaming clustering

Momentum/NAG and local minima

Hessian Free Optimization and Rectified Linear Units.

15 Votes

gradient-descent × 5
hmm × 5
torch × 5
pylearn2 × 5
pylearn × 5
libraries × 5
software × 5
algorithms × 4
dataset × 4
dropout × 4

learning × 4
datasets × 3
speech × 3
regularization × 3
mnist × 3
autoencoder × 3
speech-recognition × 3
sparse × 3
data × 3
audio × 3

machine-learning × 2
decision-trees × 2
second-order × 2
learning-rate × 2
sparse-coding × 2
random-forests × 2
hessian-free × 2
lstm × 1
backpropagation × 1
adagrad × 1

density-estimation × 1
opencv × 1
scikits-learn × 1
non-parametric × 1
kl-divergence × 1
rectified-linear-units × 1
computer-vision × 1
r × 1
nonparametric × 1
clustering × 1

20 Kudos

● Prominent Question × 15
● Well-Known Question × 15
● Conspicuous Question × 15
● Prized Question × 5
● Appreciated Question × 3
● Appreciated Answer × 2

● Esteemed Answer × 2
● Cherished Question × 2
● Esteemed Question × 2
● Participant × 1
● Mentor × 1
● Revisionist × 1

● Adulated Question × 1
● Pronouncement × 1
● Cognoscenti × 1
● Blue Pencil × 1
● Guide × 1
● Discriminating × 1

● Advocate × 1
● Inquisitive × 1

Registered user

15 Questions

5 Answers

15 Votes

50 Tags

20 Kudos