Questions tagged [representation]
The representation tag has no usage guidance.
41
questions
0
votes
0
answers
21
views
How do I represent multiple seconds of video in tensors?
I want to use ResNet-18 for image features at the head of a VAE. If I have 10 seconds of video at 30 frames per second, and if ResNet-18 gives me the features of a single frame as a 512-dimensional ...
1
vote
1
answer
19
views
Is n-gram a special instance of bag of word? What are their differences?
Is n-gram a special instance of bag of word? What are their differences? From my understanding, n-gram is when replacing the words in bag of words with n-grams, and follow the same procedures to ...
- 397
1
vote
0
answers
24
views
Is normalization of word embeddings important?
I am doing actor-critic reinforcement learning for an environment that is best represented as a "bag-of-words". For this reason, I have opted to use a single body, multi-head approach for ...
0
votes
0
answers
27
views
Using large CNNs (e.g., ResNet) in convolutional autoencoders for image representation learning
I am confused about which CNNs are generally used inside autoencoder architectures for learning image representations. Is it more common to use a large existing network like ResNet or VGG, or do most ...
- 91
0
votes
0
answers
9
views
Is Self-Supervised Learning a task of Representation Learning?
Maybe a weird question but: Currently, I'm writing a seminar paper about Self Supervised Learning for time series data. For this paper, I have to find methods to prepare unlabelled time series data ...
- 1
1
vote
1
answer
16
views
Does feature engineering require absolute accuracy?
Sometimes when I'm studying the datasets, the text field is particularly challenging to handle. For whatever features I want to derive from the text fields, I try to apply some heuristic to ...
- 397
1
vote
0
answers
37
views
Why are wavelet transforms not scale-equivariant?
One can rely on continuous wavelets to build a multi-resolution analysis that is equivariant ("covariant") under the action of a discrete subgroup of translation.
When not downsampled,
the ...
- 111
0
votes
1
answer
17
views
Output representation for a neural network to learn grid-based game with multiple units
I have a round based game played on a grid map with multiple units that I would like to control in some fashion using neural network (NN). All of the units are moved at once. Each unit can move in any ...
- 101
1
vote
1
answer
1k
views
How can i get the vector of word using BERT?
I need to get word-vectors using BERT and got this function that i think it should be the one i need
...
- 113
0
votes
1
answer
27
views
How to represent a "switch"-like behavior in a neural network?
I have three input variables $x_1$, $x_2$ and $d$, where $x_1$ and $x_2$ are numerical variables and $d$ is a dummy variable that takes the value of 1 or 2. How to represent the part of a neural ...
- 101
0
votes
2
answers
3k
views
How to JUST represent words as embeddings by pretrained BERT?
I don't have enough data (i.e. I don't have enough texts) --- have only around 4k words in my dictionary. I need to compare given words, then I need to representate it as embedding.
After the ...
- 137
2
votes
2
answers
198
views
KNN efficient implementation
The KNN algorithm is very handy and particularly suited to some of my problems, but I can't find any resources on how to implement it in production.
As a comparative example, when I use a neural ...
- 21
2
votes
1
answer
859
views
Difference between NCE-Loss and InfoNCE-Loss
I started looking into word2vec and was wondering what the connection/difference between the NCE-Loss and the infoNCE-Loss is. I get the basic idea of both.
I have a hard time deriving one from ...
- 21
3
votes
1
answer
3k
views
What are latent representations?
I am reading some research papers about graph convolutional neural networks and I have seen the term "latent representations" used a lot. For instance, "the model was able to learn ...
- 31
0
votes
1
answer
215
views
How to reconstruct a scikit-learn predictor for Gradient Boosting Regressor?
I would like to train my datasets in scikit-learn but export the final Gradient Boosting Regressor elsewhere so that I can make predictions directly on another ...
- 221
2
votes
1
answer
126
views
What are the Most Dissimilar MNIST Digits?
Using whatever definition of dissimilarity over sets that you'd like, what are the most dissimilar two digits in MNIST? I was thinking that a reasonable approach to answering the question would be to ...
- 121
1
vote
2
answers
866
views
Using categorical and continuous variables in Deep Learning
I would like to apply a MLP to some business seller data. I found that the data is a mix of both categorical and continuous features. For what I read it is not advisable to feed a neural network with ...
- 217
1
vote
1
answer
73
views
Good chromosome representation in a VRPTW genetic algorithm
I have a genetic algorithm for a vehicle routing problem with time windows and I need to implement certain modifications. I am not sure what would be the best chromosome representations.
I have tasks ...
1
vote
1
answer
388
views
Clustering with categorical as well as numerical features
I have dataset consisting of house prices for example. The dataset contains features such as: house size, monthly rent, house colour, location, year the house was built.
I wanted to group these all ...
- 11
1
vote
0
answers
32
views
Representation sample size- n [closed]
Need help with identifying a representation sample size 'n'. Let's say I have a very large population- infinite number of participants. I am picking the random sample from this infinite population. I ...
4
votes
1
answer
68
views
What does it exactly mean when we say that PCA and LDA are linear methods of learning data representation?
I have been reading on representation learning and I have come across this idea that PCA and LDA are linear methods of data representation, however, auto-encoders provide a non-linear way. Does this ...
- 307
0
votes
1
answer
80
views
A way to init sentence embedding for unsupervised text clustering, better than glove wordvec?
For unsupervised text clustering, the key thing is the init embedding for text.
If we want to use deepcluster for text, the problem for text is how to get the init embedding from deep model.
BERT can ...
- 631
2
votes
2
answers
161
views
Is it possible to compress a sequence of numbers through an autoencoder?
Specifically: I would like compress a set of coordinates, which map to the locations of 1's in a binary image, and then decode back to the original set. For instance, for a 16x16 image, the input ...
- 175
0
votes
1
answer
485
views
Embedding of list of objects
I have a dataset where each sample is a list of ordered items, lets say grocery list , and a label from 6 categories . each list can have up to 120 items but the mean items is 12 items in a list.
i ...
- 283
1
vote
1
answer
486
views
What are the sparse and dense vector ? I cant undestand ,can you explain to me?please.Why do we use for?
I am new to neural networks, embeddings, etc. I am struggling understanding things like sparse representation, embeddings, and especially sparse vectors. Could you explain these to me? Why do we need ...
- 11
0
votes
1
answer
30
views
Representing user information
I have a task of representing a users feature matrix , i have features like gender , age etc but I also have a multivalue feature called as "movies watched" which is essentially another table of ...
- 191
1
vote
0
answers
12
views
Standard(s) for data representing measurement times with their interval of validity?
Is there a standard for representing the time interval of a measurement?
Any actual measurement is made at a point in time with, perhaps, an interval over which it is considered valid. For example, ...
- 111
1
vote
2
answers
25
views
Difference between n-dimensional of length m and m-dimensional of length n?
With respect to classifying the MNIST digits I've seen the input representation referred to as both a one-dimensional array of 748 elements and a 784-dimensional vector.
Are these perspectives always ...
- 111
0
votes
1
answer
343
views
PCA, why variance of eigen values is measure of its utility?
Source - Murphy, 12.3
Heuristic for assessing applicability of PCA.
Let the empirical covariance matrix Σ have eigenvalues λ1≥λ2≥···≥λd>0, with mean λ. Explain why the variance of the eigen values, ...
9
votes
2
answers
320
views
Are there any graph embedding algorithms like this already?
I wrote an algorithm for generating node embeddings based on the graph's topology. Most of the explanation is done in the readme file and the examples.
The question is:
Am I reinventing the wheel?
...
- 201
0
votes
1
answer
14
views
Membership ratio graph
Does anyone know how the what the kind of graph in this image is called?
Each color represents a class and the height of the color, in a particular instant, represents the ratio of elements that ...
- 103
0
votes
2
answers
50
views
Is there any option to represent 160 discrete variables each one with their own color in scatter plot?
I want to know if it is possible to represent with different colours in a scatter plot more than 10 parameters instead of putting others and assigning a colour if the number of parameters overpasses ...
- 11
5
votes
1
answer
626
views
What does 'Linear regularities among words' mean?
Context: In the paper "Efficient Estimation of Word Representations in Vector Space" by T. Mikolov et al., the authors make use of the phrase: 'Linear regularities among words'.
What does that mean ...
- 8,136
1
vote
0
answers
24
views
Daily-level variables and weekly-level variables at the same model
I am running a model which has NPS (Net Promoter Score) as its target and various predictors among which some CSAT (Customer Satisfaction) data.
The rest of the predictors are good to use at daily ...
- 1,027
4
votes
1
answer
311
views
Why ELMo's word embedding can represent the word better than glove?
I have read the code of ELMo.
Based on my understanding, ELMo first init an word embedding matrix A for all the word and then add LSTM ...
- 631
1
vote
1
answer
488
views
How to deal with Optional Input
I'm from the vision world and only worked with pixels from 0-255, ignoring any side effects. My current problem is different, in the way that I cannot rely on the input data.
What my problem is:
I ...
- 13
1
vote
1
answer
807
views
faster alternatives to sparse.model.matrix?
I have a large dataset that is entirely categorical. I'm trying to train with it using xgboost, so I must first convert this categorical data to numerical. So far I've been using sparse.model.matrix() ...
- 11
2
votes
1
answer
72
views
Learning Football Player Stats like FIFA's by only the game result [closed]
It is a general question on how to learning representation of one entity but the dataset is mixed with a lot of other entities, which their statis are always waiting to be learnt.
The question is ...
- 121
1
vote
1
answer
838
views
Neural Nets: unordered sets of ordered tuples as features of data
I'm working on a very small scale pet project in which inputs are essentially sets of (x, y) pairs, and are to be classified into categories, using deep learning, specifically using Keras (I know this ...
- 31
0
votes
1
answer
574
views
How to learn from multiple data sources with different input variables but the same underlying pattern?
I will explain with an example:
Let's say you have 2 factories that produce pulp paper. Each have similar processes where the laws of physics give the same outcome. Now let's say this 2 factories ...
- 103
8
votes
2
answers
19k
views
Why don't tree ensembles require one-hot-encoding?
I know that models such as random forest and boosted trees don't require one-hot encoding for predictor levels, but I don't really get why. If the tree is making a split in the feature space, then isn'...
- 83