Skip to main content
3 votes

Why is it not always possible to compute the centroid of feature vectors?

A metric space consists of a set $X$ of "points" and a metric $d\colon X \times X \to \mathbb{R}_{\geq 0}$ (giving the "distance" between any two points) which satisfies the following constraints: ...
Yuval Filmus's user avatar
2 votes
Accepted

Why do AlphaGo and AlphaGo Zero include board history in the input features

"This doesn't seem to be discussed in either of the papers, " Yes, it is discussed in at least one of the papers. Here is an excerpt taken from Mastering the Game of Go without Human Knowledge - ...
John L.'s user avatar
  • 39k
2 votes

Why do AlphaGo and AlphaGo Zero include board history in the input features

The primary reason for including a history of states is likely indeed the ko rule. Even if having a long history will often be redundant, it's unlikely to hurt either (except that it might take some ...
Dennis Soemers's user avatar
2 votes

Use of Activation Units in Facial Expression Understanding

As a preliminary answer, I can't imagine the extraction of AUs affecting their ability to detect emotions, since according to Do Deep Neural Networks Learn Facial Action Units When Doing Expression ...
Seanny123's user avatar
  • 641
2 votes
Accepted

Classification training data, but regression prediction

This is still a binary classification task. In the abstract, there are two ways to handle this: Most classifiers can output a predicted class and a confidence score (which indicates how confident ...
D.W.'s user avatar
  • 162k
2 votes
Accepted

Largest isomorphic subgraphs of two graphs with features

This looks like it is at least as hard as the induced subgraph isomorphism problem. Suppose that all $w$ vectors are 0. Then your question is identical to asking whether for a maximal induced ...
D.W.'s user avatar
  • 162k
1 vote
Accepted

How can a Machine Learning model predict this classification problem?

Your premise is wrong. You are assuming we don't know how to compute the features on new sentences. But we do know. We construct features so that we can compute the features on any sentence ...
D.W.'s user avatar
  • 162k
1 vote

Featurizing images of different dimensions

Option 1: You can crop the images to the smallest sizes in all dimensions. However, blindly cropping images will cause you to lose important information, if you don't have a region of interest. For ...
ilke444's user avatar
  • 507
1 vote

Terminology: Difference between decision variables, features and attributes in this paper?

Feature and attribute have both the same meaning. Decision variable is the variable which is used to make a split in a decision tree in tree based algorithms, i.e. the variable on which the decision ...
89f3a1c's user avatar
  • 123
1 vote

What is the official name of a specific type of combination algorithm

You are probably looking for subsets of a set. If a set has $n$ elements, it has $2^n$ subsets when counting also the empty set (there is a clear correspondence to all bit strings of length $n$). So ...
Juho's user avatar
  • 22.6k
1 vote

Problem with Understanding "correlated attributes into a set of values of uncorrelated attributes" in PCA

It looks like you are using RapidMiner documentation as a medium to study. If that is case, then the best answer to your question and probably to some other questions of yours should be, I believe, a ...
John L.'s user avatar
  • 39k
1 vote

Computer vision methods without "pre-training"

Yup, there are lots of them. For example, image segmentation through graph cuts, camera calibration, image morphing, image stitching, reconstruction of a 3D scene from multiple 2D images, optical ...
D.W.'s user avatar
  • 162k
1 vote
Accepted

Feature selections : Still a high dimentionality

There seems to be a misconception. You can use $k$-nearest neighbors with any number of dimensions. There is no prohibition on using it with a large number of features. And in some cases $k-NN will ...
D.W.'s user avatar
  • 162k
1 vote

Computation of normalized first derivative in discrete case

As you have noticed, the denominators in the definition of $x'_t,y'_t$ ultimately have no effect, as they will be normalized away. Nonetheless, they do have significance. This definition ensures ...
D.W.'s user avatar
  • 162k
1 vote

Computation of normalized first derivative in discrete case

For the first part of your question, the significance is just to guarantee that the vector $(\hat{x}'_t,\hat{y}'_t)$ has length $1$. This means that it tells us in which direction $(x_t,y_t)$ is ...
David Richerby's user avatar
1 vote

Description of shape in a vector form

There's lots of work in the computer vision community on "shape descriptors". This might or might not be useful. See e.g., https://en.wikipedia.org/wiki/Shape_analysis_(digital_geometry), https://en....
D.W.'s user avatar
  • 162k
1 vote

How to detect Farkas or MPEG4 FDP points on image with a face?

In the simplified special case where each image $x$ has a single correct labelling $k=\kappa(x)$, this becomes just a regression problem: you want to find a function $q$ that minimizes the risk $$\...
D.W.'s user avatar
  • 162k
1 vote
Accepted

How does an image classifier select relevant features?

There is no single answer. A cascade is a very simple idea: it basically represents a bunch of classifiers, applied sequentially. You are free to decide how each individual cascade will work. You ...
D.W.'s user avatar
  • 162k

Only top scored, non community-wiki answers of a minimum length are eligible