Questions tagged [dimensionality-reduction]

Dimensionality reduction refers to techniques for reducing many variables into a smaller number while keeping as much information as possible. One prominent method is [tag pca]

Filter by
Sorted by
Tagged with
0 votes
0 answers
7 views

Clustering by using Locality sensitive hashing *after* Random projection

It is well known that Random Projection (RP) is tightly linked to Locality Sensitive Hashing (LSH). My goal is to cluster a large number of points lying in a d-dimensional Euclidean space, where $d$ ...
0 votes
0 answers
11 views

Movement Analysis (Unsupervised Learning)

I have some data on the movement of a drone, D piloted by multiple people, P. Multiple metrics on the its motion is recorded, example speed, acceleration, elevation, angles which will be denoted by X1,...
  • 13
0 votes
0 answers
23 views

Is it possible to use text Auto Encoders without text generation?

I have a use case where I have large texts, and a lot of it. Pretty often the sequence length exceeds 1000 tokens. I need a lower dimensional compression of the texts as an input for a classifier. The ...
  • 101
0 votes
0 answers
12 views

Adjusting the parameters for Isomap and Spectral Embedding/Eigenmap in sklearn?

I read around the basic ideas of these reduction methods, but I have no idea on how I would adjust the parameters for these so that it still reflects the original data as accuratley as possible. What ...
  • 11
0 votes
0 answers
12 views

Generative Autoencoder with latent vector size as a parameter?

I am interested in using a generative autoencoder (something like a VAE maybe) to sample very high dimensional data more easily (making use of the fact that the autoencoder reduces the dimensionality ...
0 votes
0 answers
20 views

Extrinsic evaluation of BERT embeddings

I have finetuned a BERT-based LM (DistilBERT Uncased) on my dataset for the purpose of sequence classification. Now I am trying to investigate the embedding of the ...
0 votes
0 answers
19 views

What prevents dimensional reduction methods from being generative?

The obvious exception in my mind would be autoencoders, but I am wondering for methods like TSNE, UMAP, and self-organizing maps, what prevents someone from generating new points with these methods. ...
0 votes
0 answers
13 views

Reducing the dimensionality of entire dataset

I have a lot of datasets with different shapes. For example few of them are (90, 892), (74, 853), (93, 765), ... etc. I want to convert this shape to (90, 4), (74, 4), (93, 4) ... (x, 4). And, after ...
1 vote
1 answer
35 views

How to describe this UMAP connectivity figure

I generated this UMAP connectivity diagram of my research data. How do I interpret/describe this plot regarding the UMAP connectivity? Is it correct to say that: As there is a lot of connectivity ...
  • 135
0 votes
0 answers
26 views

How to use word embeddings and numerical features together for UMAP visualization?

I'm trying to visualize a large collection of artwork with both numerical and text features (e.g., its physical dimensions and text description), and I want to incorporate all these features together ...
  • 1
0 votes
1 answer
38 views

Classification problem with no context in numerical features

I have an extremely abstract and numeric data with equally abstract objective. I have around 3000 rows of train data (df_train), where I have a binary target ...
  • 118
2 votes
0 answers
19 views

How to understand to what maximum size you can reduce the dimension of data and avoid the curse of the dimensionality?

i have a question, maybe someone could help me. I use t-sne (also tried umap) to reduce the dimensionality of the text embeddings dataset (size of embedding 300). after that I will cluster using ...
  • 163
0 votes
0 answers
27 views

How can "profiling" be done in a dataset? What are the different techniques?

I am currently doing an analysis in which I need to "profile" each record. For example, let's say I have a dataset of accounts with customer information (name, id, address, money spent, ...
0 votes
0 answers
9 views

Best Way to Display Multidimensional Data in Tables and Graphs for a Specific Example

I am currently struggling with displaying data that I gathered in a proper data scientific way. About the Data For a Uni project I have to create different implementations of the creation of a Voronoi ...
0 votes
0 answers
15 views

PCA, Better performances with 300 components rather than 400 components : why?

I am building this content based image retrieval system. I basically extract feature maps of size 1024x1x1 using any backbone. I then proceed to apply PCA on the extracted features in order to ...
1 vote
1 answer
45 views

What is the Purpose of Feature Selection

I have a small medical dataset (200 samples) that contains only 6 cases of the condition I am trying to predict using machine learning. So far, the dataset is not proving useful for predicting the ...
  • 345
0 votes
0 answers
16 views

How to Approach Linear Machine-Learning Model When Input Variables are Inconsistent

Disclaimer: I'm relatively new to the data science and ML world -- still trying to get a firm grasp on the fundamentals. I'm trying to overcome a regression challenge involving a large, multi-...
0 votes
1 answer
43 views

An autoencoder setup for anomaly detection

I am doing anomaly detection using machine learning. i have tried different models like isolation forest, SVM and KNN. The maximum accuracy that I can get from each of them is $80\%$ accordind to my ...
  • 11
-1 votes
2 answers
144 views

Why Do a Set of 3 Clusters Across 1 Dimension and a Set of 3 Clusters Across 2 Dimensions Form 9 Apparent Clusters in 3 Dimensions?

I am sorry if this is a well-known phenomenon but I can't quite wrap my head around this. I have a related question: How To Develop Cluster Models Where the Clusters Occur Along Subsets of Dimensions ...
1 vote
0 answers
36 views

Accuracy drops when adding a fully connected layer for dimensionality reduction to a ResNet50

I'm training a ResNet50 for image classification and I'm interested in decreasing the dimensionality of the embedded layer, in order to apply some clustering techniques. The suggested dimension is ...
  • 61
0 votes
0 answers
37 views

Would you ever chose t-SNE over UMAP?

UMAP is both faster and captures the global structure better than t-SNE when visualizing high-dimensional data. Is there ever a situation where you would pick t-SNE over UMAP?
0 votes
0 answers
14 views

Single scalar from vector

I am aware that this question is very general, but I found this question and it made me curious. What are the sensible ways that you can think of to derive a single scalar value from a vector? Of ...
  • 21
0 votes
1 answer
18 views

Dimensionality Reduction of Curved Structural Data

I have been using PCA dimensionality reduction on datasets that are quite linear and now I am tasked with the same on datasets that are largely curved in space. Imagine a noisy sine wave for ...
  • 111
0 votes
0 answers
18 views

Clustering Algorithm + Euclidean Distance to find similarities

Goal: Create a tool that recommends similar players based on their statistical profile Process: (1) Standardize data (2) UMAP to reduce dimensionality (c. 50 features) (3) First-Stage Clustering: GMM ...
1 vote
1 answer
16 views

How to choose Recursive Feature Elimination parameters

in my project I have >900 features and I thought to use Recursive Feature Elimination algorithm to reduce the dimensionality of my problem (in order to improve the accuracy). But I can't figure out ...
0 votes
0 answers
20 views

Searching machine learning algorithm for regression problem with many features

I have a machine learning problem with about 160 features and 400 cases and I want to find the best predictors for a continuous outcome. The dataset contains variables of psychotherapists and clients. ...
2 votes
1 answer
194 views

What is the meaning of preserving local or global structure of the data?

I read about PaCMAP dimensionality reduction method (PaCMAP). They wrote that this method preserving both local and global structure of the data in original space. ...
  • 277
1 vote
2 answers
66 views

Using PCA for Dimensionality Expansion

I was trying to use t-SNE algorithm for dimensionality reduction and I know this was not the primary usage of this algorithm and not recommended. I saw an implementation here. I am not convinced about ...
  • 13
0 votes
0 answers
37 views

PCA in supervised machine learning

I would like to use PCA in supervised ML aiming to generate a binary classification model. The data set I have consists of well validated target variable (labels) concerning the one classification ...
2 votes
0 answers
147 views

Can one use PCA to reduce the dimensionality of One-Hot-Encoded data?

I read a couple times that PCA was used as a method to reduce dimensionality for one-hot-encoded data. However, there were also some comments that using PCA is not a good idea since one-hot-encoded ...
0 votes
0 answers
11 views

Feature Selection for vector of targets, Y=(y1,...,yn)

If I have an output/target, y, that is a vector of targets, e.g. (y1(t), y2(t), y3(t), y4(t)) and training data X, also a vector, (x1(t),...,xn(t)), and I wish to do e.g. regression, neural nets, and ...
  • 71
1 vote
0 answers
13 views

Nearest neighbor face recognition in eigenspace when using dot product of test set with eigenvectors does not match the performance when using sklearn

I am trying to perform Face recognition using PCA (eigenfaces). I have a set of N training images (of dimensions M=wxh), which I have pre-processed into a vertical ...
0 votes
0 answers
13 views

Practical significance of number of singular values in SVD

I am working on a binary classification problem. SVD is used for dimensionality reduction and the vector with reduced dimension is used as the feature vector. DNN is used as the classifier. There are ...
1 vote
0 answers
23 views

Understanding Linear Discriminant Analysis

I am trying to understand the results of a Linear Discriminant Analysis. For that purpose I used the Iris dataset. I plotted the two first features: Then, I projected these two features in the new ...
0 votes
0 answers
15 views

When should I use dimensionality reduction in cluster analysis

I would like to ask a question about dimensionality reduction in cluster analysis. I currently have a table data with 15 rows and 120 columns, and I'll doing cluster analysis on this data. In this ...
  • 1
0 votes
0 answers
18 views

How to reduce columns with binary data

I have one-hot encoded some categorical columns and as you might have guessed the resulting data frame is very large and I'm looking for a way to reduce the dimensions of this data frame. PCA is not ...
1 vote
2 answers
76 views

Dimensionality reduction for millions of features

I have a dataset with 10 million observations and 1 million sparse features. I would like to build a binary classifier for predicting a particular feature of interest. My main problem is how to deal ...
  • 11
0 votes
0 answers
34 views

Isomap/LLE python library that allows precomputed distance matrix

I want apply some algorithm to the distance matrix before reducing dimensionality using Isomap, LLE and Laplacian Eigenmaps. Are there Python libraries I can use for this?
2 votes
1 answer
654 views

Can t-SNE be applied to visualize time series datasets

I have multiple time-series datasets containing 9 IMU sensor features. Suppose I use the sliding window method to split all these data into samples with the sequence length of 100, i.e. the dimension ...
0 votes
0 answers
19 views

Projection of PCA into new variable space (study area)

I have a set of >30 variables distributed in x,y coordinates across two study areas. I'd like to reduce that dimensionality with a PCA applied in the first study area. Then I'd like to visualize ...
0 votes
0 answers
17 views

Best way to classify a high dimension dataset

I am writing a code to classify 4-body systems as 'stable' and 'unstable'. For this purpose, I generated a vast (~1 million) dataset of 4-body systems, and narrowed down the number of possible ...
  • 53
0 votes
0 answers
134 views

Variational autoencoder for time series denoising and dimentionality reduction

I have a dataset X of multiple series say 100 (size=100). I would like to use VAE to both denoise the data and reduce the dimensions to a smaller latent space Z (size Z << size X), because I ...
0 votes
0 answers
7 views

How to interpret data projected on the sum of first few principal components weighted by eigen values?

I have simulation time series data of a molecule from Molecular dynamics and I want to visualize the very high-dimensional trajectory in two dimensions and also identify some clusters. The problem is ...
  • 1
0 votes
0 answers
56 views

PCA vs t-SNE in asset pricing

So I am trying dimensionality reduction techniques on the S&P500 FY2020 data. I understand the CAPM model and the fact that doing a PCA determines my market variability factor (the first PCA ...
1 vote
1 answer
81 views

Factor Analysis vs PCA

Could someone please explain when FA is used or when PCA is used, as I understood FA do dimensionality reduction, however PCA - the main goal is the same. Then which one should I use and in which ...
  • 33
0 votes
0 answers
14 views

How to classify (supervised) a multi dimensional vector?

What kinds of machine learning tool is used to classify a vector of data which are not spatially correlated? I have a 158*158 image*15000 samples which I tried to ...
1 vote
1 answer
135 views

Reusing UMAP transformation

I have this use case: I want to apply a dimension reduction with UMAP to a initial dataset of high-dimension vectors (100d), and later, in second place, have the oppurtinity to add new data points ...
1 vote
0 answers
543 views

What are the differences among Proper Orthogonal Decomposition (POD), Singular value decomposition (SVD) and principal component analysis (PCA)?

Proper Orthogonal Decomposition (POD), singular value decomposition (SVD), and principal component analysis (PCA) are three eigenvalue methods used to reduce a high-dimensional data set into fewer ...
0 votes
0 answers
58 views

How to implement the contribution analysis using PCA?

I have been looking into implementing the Q-Residual and Hotelling's T statistics calculation to the PCA components which is similar to the following article and website: Structural Health Monitoring ...
0 votes
1 answer
39 views

Latent Semantic Analysis on image features using k means

I have extracted Color Moments from a set of images and want to use Kmeans to perform dimensionality reduction and find the top k latent semantics. How can I use Kmeans for latent semantic analysis?

1
2 3 4 5 6