16
votes
How can I encode angle data to train neural networks?
The main problem with simply using the values $\alpha \in [0, 2\pi]$ is that semantically $0 = 2\pi$, but numerically $0$ and $2\pi$ are maximally far apart. A common way to encode this is by a vector ...
- 1,317
5
votes
Accepted
Is there an argument against using the (reviewed) predictions of a model as ground truth to further train exactly this model?
Using the (unchecked) predictions of the model as training data is an approach known as "pseudo-labeling". It can help in certain situations, depending on the underlying structure of your ...
- 278
4
votes
Accepted
Why does MNIST provide only a training and a test set and not a validation set as well?
The test set should never be seen and ran once at the end of training.
The validation set is used to help you select hyperparameters and it would be cheating to tune your model on the test set because ...
- 56
4
votes
Is there an argument against using the (reviewed) predictions of a model as ground truth to further train exactly this model?
The answer is: It depends.
What you describe is a strategy often used to save time and costs for labelling data. It is important that the data you have already labelled (the 20%) is representative of ...
- 205
3
votes
Accepted
What is the total number of actions and rewards count
TL;DR In the DQN paper, each environment was trained for 50 million frames, grouped in fours without overlap, so there were 12.5 million state, action, reward next-state records used.
The above direct ...
- 26.5k
2
votes
Is there an image classification dataset where the class depends on spatial relations?
There is no specific image classification dataset that focuses on spatial relations. However, there are some datasets that include images with spatial relations annotations, such as the Visual ...
- 1,004
2
votes
Does more data increase training accuracy in neural networks?
The question use specific terms in a vague way so let me set some very basic ground definitions first. It might sounds trivial but please bear with me cause it's easy to give reasonable answers that ...
- 4,753
2
votes
To train a mushroom vs. non-mushroom image classifier, which images should comprise the negative class?
In training a mushroom vs. non-mushroom image classifier, which images should comprise the negative class?
Only you can answer this question. The answer requires some careful thinking on your part, ...
- 882
2
votes
Accepted
Is data useless for a neural network if some inputs are derivatives of other inputs?
No it is not useless.
The relationship may not be obvious, and having the data will allow the network to learn this 𝑓 relationship.
Further, even if 𝑓 is obvious, networks are so sample inefficient ...
- 2,005
1
vote
How can imitation learning data be collected?
Imitation learning data usually means data gathered from an expert, that is data from an agent proficient in the task.
The agent may be:
A human operator: have the operator complete the task and ...
- 1,056
1
vote
Split dataset into Train/Validation/Test for Object Detection
You can use the scikit learn train_test_split() by passing the stratify argument with the class value.
- 431
1
vote
Accepted
Are the "artifacts" in select Keras MNIST training images really there or is my download corrupt?
There are many versions of MNIST digits now, and some of them are corrupted, binarized, or otherwise altered (see TensorFlow datasets) intentionally; but I don't think the keras MNIST digits has these ...
- 882
1
vote
Why does data augmentation using synthetic data generated by one model improve the performance of another model?
The reasoning behind synthetic data is the same behind classic data augmentation,so the goal is to increase the amount of training instances to improve generalization.
The difference with classic data ...
- 4,753
1
vote
Does more data increase training accuracy in neural networks?
Using more training samples decreases the chance of over-fitting. However, I think, it may not occasionally result in a decrease in the training error, maybe the opposite (look at the loss function ...
- 69
1
vote
Feature Engineering on transactional dataset clustering
The average transaction is a central measure, while the minimum and maximum transactions together give an idea of dispersion. However, these can be very sensitive to individual purchases that might ...
1
vote
Accepted
Should I include overlapping (input) Data in my training data
In general, both methods are valid to train temporal models. The only thing you need to check is that validation and test-set don't overlap with any of your training samples.
Using the overlapping ...
- 1,317
1
vote
What makes a 'good' dataset
This is not a simple answer, and I think it really depends the goal, quality is not clearly defined and can vary. However there are some points that are commonly seen as positive.
The quantity
As you ...
- 61
1
vote
How to deal with an unbalanced dataset?
4861/5111 is about 95.1%, so it looks like your classifier is probably predicting every patient as "no stroke" (i.e. it is not really doing anything useful). The thing to do is to work out ...
- 121
Only top scored, non community-wiki answers of a minimum length are eligible
Related Tags
datasets × 259machine-learning × 75
neural-networks × 57
deep-learning × 44
training × 38
convolutional-neural-networks × 31
data-preprocessing × 28
classification × 21
image-recognition × 12
object-detection × 11
training-datasets × 11
natural-language-processing × 10
computer-vision × 10
reinforcement-learning × 9
object-recognition × 9
data-science × 9
terminology × 8
long-short-term-memory × 8
unsupervised-learning × 8
supervised-learning × 8
reference-request × 7
tensorflow × 7
python × 7
math × 7
data-augmentation × 7