User The Pointer

10 votes

2 answers

756 views

Can machine learning algorithms be used to differentiate between small differences in details between images?

asked Jan 14, 2019 at 5:38

7 votes

2 answers

704 views

How should we interpret this figure that relates the perceptron criterion and the hinge loss?

asked Nov 21, 2020 at 13:19

6 votes

0 answers

107 views

$\frac{P(x_1 \mid y, s = 1) \dots P(x_n \mid y, s = 1) P(y \mid s = 1)}{P(x \mid s = 1)}$ indicates that naive Bayes learners are global learners?

asked Dec 13, 2020 at 22:15

5 votes

1 answer

152 views

Many of the best probabilistic models represent probability distributions only implicitly

asked Feb 25, 2020 at 8:09

3 votes

1 answer

110 views

Research paths/areas for improving the performance of CNNs when faced with limited data

asked Sep 15, 2020 at 11:59

3 votes

2 answers

384 views

Can we apply transfer learning between any two different CNN architectures?

asked Sep 20, 2020 at 16:30

3 votes

1 answer

86 views

What is meant by "the number of examples is reduced", and why is this the case?

asked Dec 19, 2020 at 13:46

3 votes

1 answer

88 views

Is this referring to the true underlying distribution, or the distribution of our sample?

asked Nov 25, 2020 at 5:40

3 votes

1 answer

140 views

In the machine learning literature, what does it mean to say that something is "embedded" in some space?

asked Dec 30, 2020 at 17:23

3 votes

1 answer

408 views

What is meant by "stable training" of a deep learning model?

asked May 30, 2021 at 18:03

2 votes

1 answer

569 views

What is an "input embedding" in the context of NLP?

asked May 30, 2021 at 20:31

2 votes

0 answers

61 views

How does the support vector machine constraint imply that sample selection bias will not systematically affect the output of the optimisation?

asked Dec 25, 2020 at 12:05

2 votes

1 answer

177 views

Should I remove the text overlaying some images in the dataset before training the CNN?

asked Sep 17, 2020 at 0:44

2 votes

1 answer

705 views

How do sigmoid functions make it so that the prediction $\hat{y}$ indicates the probability that the observed value, $y$, is $1$?

asked Jun 11, 2021 at 10:52

2 votes

3 answers

333 views

Where does the so-called 'loss' / 'loss function' fit into the idea of a perceptron / artificial neuron (as presented in the figure)?

asked Jun 13, 2021 at 23:36

2 votes

1 answer

117 views

Why doesn't the set $\{ -2, +2 \}$ in $E(X) = (y − \text{sign}\{\overline{W} \cdot \overline{X} \}) \in \{ −2, +2 \}$ include $0$?

asked Nov 21, 2020 at 13:08

1 vote

1 answer

73 views

Why is ancestral sampling used in autoregressive models?

asked Jun 24, 2021 at 11:29

1 vote

1 answer

185 views

An explanation involving the sign activation, its affect on the loss function, and the perceptron and perceptron criterion: what is this saying?

asked Jun 25, 2021 at 11:16

1 vote

1 answer

375 views

What is a "continuous vector"?

asked Jul 4, 2022 at 11:50

1 vote

1 answer

94 views

Can most of the basic machine learning models be easily represented as simple neural network architectures?

asked Nov 7, 2020 at 16:51

1 vote

1 answer

179 views

Solving the supervised learning problem of learning $p(y \vert \mathbf{x})$ by using traditional unsupervised technologies to learn $p(\mathbf{x}, y)$

asked Feb 25, 2020 at 10:40

1 vote

2 answers

217 views

What does "the expectation is taken across different possible inputs, drawn from the distribution of inputs we expect the system to encounter" mean?

asked Mar 2, 2020 at 10:51

1 vote

2 answers

216 views

Why does the machine learning algorithm need to learn a set of functions in the case of missing data?

asked Feb 20, 2020 at 8:34

1 vote

2 answers

100 views

What does 'clock rate' mean in the context of recurrent neural networks (RNNs)?

asked Apr 7, 2021 at 14:00

0 votes

1 answer

46 views

What is it about sigmoid activations in particular that allows for the keeping and forgetting of past information from different time scales?

asked Jun 4, 2021 at 18:43

0 votes

1 answer

60 views

What would the "state space" and its Python implementation be for my simulation?

asked Oct 13, 2022 at 9:54

Stack Exchange Network

26 Questions

Can machine learning algorithms be used to differentiate between small differences in details between images?

How should we interpret this figure that relates the perceptron criterion and the hinge loss?

$\frac{P(x_1 \mid y, s = 1) \dots P(x_n \mid y, s = 1) P(y \mid s = 1)}{P(x \mid s = 1)}$ indicates that naive Bayes learners are global learners?

Many of the best probabilistic models represent probability distributions only implicitly

Research paths/areas for improving the performance of CNNs when faced with limited data

Can we apply transfer learning between any two different CNN architectures?

What is meant by "the number of examples is reduced", and why is this the case?

Is this referring to the true underlying distribution, or the distribution of our sample?

In the machine learning literature, what does it mean to say that something is "embedded" in some space?

What is meant by "stable training" of a deep learning model?

What is an "input embedding" in the context of NLP?

How does the support vector machine constraint imply that sample selection bias will not systematically affect the output of the optimisation?

Should I remove the text overlaying some images in the dataset before training the CNN?

How do sigmoid functions make it so that the prediction $\hat{y}$ indicates the probability that the observed value, $y$, is $1$?

Where does the so-called 'loss' / 'loss function' fit into the idea of a perceptron / artificial neuron (as presented in the figure)?

Why doesn't the set $\{ -2, +2 \}$ in $E(X) = (y − \text{sign}\{\overline{W} \cdot \overline{X} \}) \in \{ −2, +2 \}$ include $0$?

Why is ancestral sampling used in autoregressive models?

An explanation involving the sign activation, its affect on the loss function, and the perceptron and perceptron criterion: what is this saying?

What is a "continuous vector"?

Can most of the basic machine learning models be easily represented as simple neural network architectures?

Solving the supervised learning problem of learning $p(y \vert \mathbf{x})$ by using traditional unsupervised technologies to learn $p(\mathbf{x}, y)$

What does "the expectation is taken across different possible inputs, drawn from the distribution of inputs we expect the system to encounter" mean?

Why does the machine learning algorithm need to learn a set of functions in the case of missing data?

What does 'clock rate' mean in the context of recurrent neural networks (RNNs)?

What is it about sigmoid activations in particular that allows for the keeping and forgetting of past information from different time scales?

What would the "state space" and its Python implementation be for my simulation?