Artificial Intelligence Stack Exchange
Artificial Intelligence Stack Exchange

Questions tagged [dimensionality]

13 questions
3
votes
2 answers

Problem extracting features from convolutional layer where the dimensions are big for feature maps

I have trained a convolutional neural network on images to detect emotions. Now I need to use the same network to extract features from the images and use them to train an LSTM. The problem is: the dimensions of the top layers are: [None, 4, 4, 512]…
I. A
  • 131
  • 3
2
votes
1 answer

Is there any advantage to providing multi-dimensional input to torch modules?

Most layer types in torch.nn such as torch.nn.Linear accept input with more than one dimension. Is there any advantage in doing so if you can shape your data to represent a certain arrangement in order to encode positional information? For example,…
kot
  • 23
  • 2
1
vote
1 answer

curse of dimensionality v.s. volume of n-dimensional ball v.s. n-dimensional vector embedding space

We know that the performance of machine learning model become worse if we feed the model with a few features and many features (high dimensional data). This is known as the curse of dimensionality. The relationship between performance and the num of…
Muhammad Ikhwan Perwira
  • 800
  • 3
  • 10
1
vote
1 answer

What is the channel dimension other than color representation in Conv2D? Shall I use Conv3D instead?

I have grayscaled image, specifically medical data ultrasonography. In the context of medical domain, there are techniques to capture data that called "View". It's like point of views of CCTVs with different placement when capturing image. So, there…
Muhammad Ikhwan Perwira
  • 800
  • 3
  • 10
1
vote
1 answer

Can I do state space quantization using a KMeans-like algorithm instead of range buckets?

Are there any reference papers where it is used a KMeans-like algorithm in state space quantization in Reinforcement Learning instead of range buckets?
ddaedalus
  • 947
  • 1
  • 7
  • 21
1
vote
0 answers

How to handle a high dimensional video (large number of frames per video) data for training a video classification network

I have a video dataset as follows. Dataset size: 1k videos Frames per video: 4k (average) and 8k (maximum) Labels: Each video has one label. So the size of my input will be (N, 8000, 64, 64, 3) 64 is height and width of video. I use keras. I am…
manv
  • 11
  • 2
0
votes
0 answers

How to tell a model that there are another useful features?

Suppose there is problem of image detection in a CCTV. Look these tensor dimensions below: Input: Image -> (Height, Width, RGB) Output: Mask of Image -> (Height, Width, Num of Object Class) object class for example person, vehicle, and others…
Muhammad Ikhwan Perwira
  • 800
  • 3
  • 10
0
votes
0 answers

Do neural networks have a perception of space, regardless of dimensionality?

Suppose I have a model M which outputs a three-dimensional tensor of size 3x3x3. I have another model N which outputs a one-dimensional tensor of size 27. Train both models on some arbitrary objective which requires the knowledge of space. For…
schmixi
  • 1
  • 2
0
votes
1 answer

Are LLM parameters synonymous with dimensions?

For example, would a Large Language Model (LLM) with parameter size 140 Billion have 140 Billion dimensions as defined in deep learning as the number of nodes in the input layer? Another way to ask this might be: Is 140B parameters the same as…
geominded
  • 101
0
votes
1 answer

How many directions of gradients exist for a function in higher dimensional space?

Gradients are used in optimization algorithms. Based on the values of gradients, we generally update the weights of a neural network. It is known that gradients have a direction and the direction opposite to the gradient should be considered for…
hanugm
  • 4,102
  • 3
  • 29
  • 63
0
votes
2 answers

How to visually or intuitively understand single element multi-dimensional tensors?

Consider the following code in PyTorch >>>torch.tensor([8]).shape torch.Size([1]) >>>torch.tensor([[8]]).shape torch.Size([1, 1]) >>>torch.tensor([[[8]]]).shape torch.Size([1, 1, 1]) We can notice that we want to store only a single element $8$…
hanugm
  • 4,102
  • 3
  • 29
  • 63
0
votes
1 answer

How does t-SNE preserves embedding orders?

According to the triplet loss Wikipedia page: t-SNE (t-distributed Stochastic Neighbor Embedding) preserves embedding orders via probability distributions, whereas triplet loss works directly on embedded distances. I don't understand how does…
Revolucion for Monica
  • 252
  • 1
  • 7
0
votes
0 answers

Huge dimensionality of input and output — any recommendations?

At work there is an idea of solving a problem with machine learning. I was assigned the task to have a look at this, since I'm quite good at both mathematics and programming. But I'm new to machine learning. In the problem a box would be discretized…
md2perpe
  • 1
  • 2