Questions tagged [unsupervised-learning]
Finding hidden (statistical) structure in unlabelled data, including clustering and feature extraction for dimensionality reduction.
672
questions
0
votes
0
answers
9
views
Silhouette Score for ordered clusters
My clusters are arranged according to a time series, and I want to compute the silhouette score for the clustering performed, considering that they follow an order. Therefore the nearest cluster to ...
- 101
1
vote
0
answers
21
views
Learning based on correlations?
I am currently interested in a specific supervised learning sub-problem. We have access to Data X and targets y as in traditional statistical / supervised learning. However both X and y are very noisy....
- 1,467
0
votes
0
answers
12
views
Correctly evaluating unsupervised learning model
I am trying to compare various unsupervised machine learning models to detect anomalous water consumption in each user's house. Now I have 10 datasets (minutely data, no anomalous points) that have no ...
- 101
1
vote
0
answers
23
views
Unsupervised Ranking
Assume we have N data points like
data = [(x1, y1, z1), (x2, y2, z2),....., (x_N, y_N, z_N)].
I am looking for unsupervised ranking algorithms. The thumb of rules ...
- 337
1
vote
1
answer
106
views
How to compare labels from clustering analysis and original ones?
I was asked to run a clustering analysis to assess the validity of labels for a manually labelled dataset.
I can simply save the actual labels (4 classes: 0, 1, 2, 3) and run clustering analysis (let'...
- 15
1
vote
1
answer
41
views
Unsupervised learning: How to identify differences between clusters?
I'm learning about unsupervised learning and I tried to use KMeans, AgglomerativeClustering and DBSCAN on the same datase. The result was ok, they seems to work fine according silhouette_score() ...
1
vote
1
answer
26
views
How can I draw the decision boundary for a simple competitive network?
I know how to train a simple competitive network. Let's say I have three inputs $x_1, x_2, x_3$ and learning coefficient $\eta=0.5.$ Let's say I have two neurons $w_1, w_2$. For each input I will ...
- 125
0
votes
0
answers
31
views
How normal is "normal enough"? for a distribution [when using Symbolic Aggregate ApproXimation (SAX)]
I am comparing 4000 time-series (each with 100 points). I would like to apply SAX to discretize these time-series into distinct events in order to make comparisons between/across the time-series
My ...
0
votes
0
answers
17
views
Are web-scale deep learning models considered to be supervised or unsupervised?
Consider recent large-scale deep learning models, such as transformers for NLP applications, or diffusion models for text-to-image generation. These models are trained on huge, readily-available ...
0
votes
0
answers
25
views
K-means clustering of 3D data doesn't seem group data points based on their pattern
I have some 3D data and I used python programming to cluster these data. Based on elbow plot, I decided that 2 clusters would be the best choice of the number of clusters (k).
When I did kmeans ...
- 93
1
vote
0
answers
25
views
Can K-means put most of the noise in the same cluster?
I am working on clustering text data (very short sentences) vectorized with tf-idf. The data are characterized by high sparseness and the presence of abundant noise (considered here as documents that ...
- 11
0
votes
0
answers
16
views
Boltzmann machine method of evolution
When a Boltzmann machine modeling binary data 'evolves' to a lower energy state, does it typically evolve via the hebbian update rule (like hopfield nets), or gradient descent? My understanding is ...
0
votes
0
answers
17
views
Clustering small group versus very big one
I'm working with a big set of data and I want to do some unsupervised clustering about the acoustic events that occur in a big set of recordings over a long time.
Some of this events are very ...
- 1
1
vote
1
answer
92
views
Do you use the PC1, PC2, PC3 or do you use PCA for feature selection in supervised learning? [closed]
My goal is to understand what is PCA for in supervised learning.
Do we use the PC1, PC2, PC3 to the supervised learning?
Do we use the generated labels to the supervised learning?
If we use the PC1, ...
1
vote
0
answers
22
views
Unsupervised many-to-one matching for two categorical variables
I have data on the sales of products sold at various retailers. My interest is in uniquely matching the product class to another arbitrary class.
For example, for two sets of hypothetical classes:
...
- 41
0
votes
0
answers
18
views
A/B testing for channel preference
Right now my organization runs a lot of promotional campaign every month. We send email, SMS and WhatsApp to all customers for each campaign. I am running a project to identity the best channel for a ...
0
votes
0
answers
61
views
Cycle detection on unsupervised time series data
i have some video data of production lines of some manufactories. In every video, an operator does the same 3-4 steps periodically for the entire video. Each periods of same steps is called cycle and ...
- 1
0
votes
0
answers
23
views
How to use GMM for clustering?
After running FAMD, the scatterplot of PC1 vs. PC2 looks like this:
It seems like if I want to do a clustering on this data, GMM is the best option (if not, please let me know what to use).
Using BIC,...
- 91
2
votes
2
answers
495
views
Implementing cross-validation to tune the hyperparameters of an unsupervised model
I have extensively researched the application of cross validation for unsupervised learning (as it is a requirement by my project manager) but it seems that there is no clear consensus as to how to ...
- 35
1
vote
1
answer
78
views
Does Discriminator in GAN Train only on Real Data or it also Trains on Fake Generated Data
I have been studying GANs and I got confused in the training phase from the discriminator. Which I think only trains on Real data, not on the generated data which then helps in distinguishing or ...
1
vote
0
answers
24
views
Is there an improved canonical polyadic decomposition for symmetric tensors?
Let us suppose I want to find a CP decomposition of a $n$-mode tensor $\mathcal{A}$. Fortunately the tensor has the permutation symmetry
$$\mathcal{A}[i_1, \cdots, i_n] = \mathcal{A}[\sigma (i_1), \...
- 6,115
0
votes
0
answers
30
views
Clustering by unsupervised RandomForest proximities, choosing parameters
Dataset: Clinical healthcare data from registry, N = 13'000, variables = 27 with different class types (numeric, categorical), outcome = survival (~20%) or death (~80%).
Goal: Explanation. Identifying ...
- 1
0
votes
0
answers
35
views
Required Properties of Labels for Disentangled Representation Learning
I came across this awesome paper "Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representations" (also described in this nice blog post), where the authors show ...
- 57
0
votes
0
answers
15
views
ML classifier comparison questions
I got some review comments as below from the conference panel regarding my results
They are comparing unsupervised learning (KNN, Linear Regression) to supervised learning (CNN, RF). How are they ...
2
votes
0
answers
63
views
Why use IsolationForest over other supervised methods for semi supervised learning?
I have a dataset with labels that I'm using to explore unsupervised learning (IsolationForest) with. IsolationForest has a few hyperparameters, and some can be heuristically determined like maybe you ...
- 21
1
vote
0
answers
37
views
Unsupervised learning (clustering) before supervised learning [closed]
Is it a common practice to do clustering before supervised learning to eliminate "noisy data"? Obviously, depending on the type of task. It seems like it makes sense in my case and my neural ...
- 11
0
votes
0
answers
29
views
Alternative to PCA that maintains densities/distances
The question is the same as posed in the title; is there an alternative to PCA that doesn't rely on the linear assumption but maintains distances (i.e. the main issue with UMAP/tSNE)?
Thanks!
3
votes
2
answers
35
views
What is the relationship between noise reduction and dimension reduction?
My understanding is that unsupervised methods like PCA, autoencoders and K-means shape a data space such that the modified representation of the data either nicely separates different families of data ...
- 43
0
votes
0
answers
54
views
How to evaluate the performance of recommender systems without having labeled data
I have a huge citation graph of research papers and datasets. So, there is an edge among two items when one of them cites another. So far I've used Node2Vec for creating a dataset recommender system ...
- 1
1
vote
0
answers
16
views
input data for training USAD network [closed]
During reading 'USAD: UnSupervised Anomaly Detection on Multivariate Time Series' paper, to train the autoencoder the authors only used normal datasets. But since this algorithm
is used for ...
- 11
1
vote
0
answers
42
views
Why separating the data in training and test sets is not feasible in unsupervised learning problem?
Based on my understanding:
Unsupervised learning problems are modeling data with no labels. Hence, we try to cluster a given data into clusters.
Supervised learning problems are modeling data with ...
- 1,378
0
votes
0
answers
17
views
How to interpret unsupervised results
I have applied some unsupervised techniques on text data in order to do topic modeling such as LDA. Once training is completed LDA produces associations between documents and topics. Now my question ...
- 1
0
votes
0
answers
24
views
Labeling of Unlabeled data
I have some labeled data and i another unlabeled dataset how to find top 5 label from labeled data for each data of unlabeled data.Let's says 5 clostest one.
I Have claculated the LSI of unlabeled ...
1
vote
1
answer
74
views
How to determine agreement between clustering methods?
Let's say you want to compare the outcome of KMeans and KMedoids. How to determine if cluster 1 from KMeans can be compared with cluster 1 with KMedoids. Or, in other words, let's say KMeans labels ...
0
votes
0
answers
52
views
K-Prototypes did not form clusters
I implemented a K-Prototypes algorithm (Huang) to cluster some mixed data in order to solve a customer segmentation question. There aren't a crazy amount of observations (n = ~6k) and with 8 fields (2 ...
- 103
1
vote
0
answers
76
views
KNN(k-nearest neighbor) algorithm as supervised and as unsupervised algorithm. What are the main differences? How can it be both?
On internet and in articles KNN ist mostly described as supervised algorithm. But recently I have find also few articles where it is mentioned as unsupervised algorithm.I cannot find articles that are ...
- 11
2
votes
1
answer
157
views
Unsupervised classification of objects based on relationships
I have size measurements of 1000 objects, measured over time. I would like to classify the objects based on the response of their size to time using unsupervised classification.
For example, the size ...
2
votes
1
answer
44
views
Infer limits of unscaled values from their standardized values - Clustering
I am working on a clustering problem and I have some skewed variables.
So, I log transform them and use them in clustering.
However, instead of multivariate clustering, I do multiple univariate ...
- 2,310
2
votes
1
answer
60
views
Converting unsupervised to supervised problem - Overfitting - bad?
I am working on a customer segmentation using 5 features such as recency, frequency, monetary, tenure, unique_product_cnt etc.
So, I did a RFM based segmentation where I used ...
- 2,310
1
vote
1
answer
153
views
standardization/normalization for 1D clustering?
I have two input variables revenue and age. Am trying to find different bins within that variables.
For ex: I have ...
- 2,310
1
vote
1
answer
2k
views
silhouette score vs Distortion score
I am working on segmenting my customers with clustering. My dataset size is 7315 rows and 30 features.
So, as a beginner to clustering, I passed all my 29 features (excluding id column) to the cluster....
- 2,310
2
votes
1
answer
165
views
Meaningful to retrieve original value after standardization using clustering
I already referred these posts here and here.
Currently, I am working on customer segmentation using their purchase data.
So, my data has below info for each customer
Based on the above linked posts ...
- 2,310
0
votes
0
answers
17
views
Classifying words/phrases that say the same thing in slightly different ways
I have survey data in which participants can manually write one or more skills that they possess. Many responses are similar enough to be classified as the same but are either not identical (e.g., &...
- 11
1
vote
1
answer
116
views
RFM Customer segmentation - Why Avg monetary value instead of total monetary value?
I am trying to segment our customers based on their purchase data. And I came to know about the RFM technique (Recency, Frequency and Monetary) through these posts here, here etc.
Recency - How ...
- 2,310
2
votes
1
answer
51
views
Image Clustering (Unsupervised learning) on unknow class(guess less than 300)
I have 30000 unlabeled images (each image has only one character), and the content of the images is very simple, basically black lines(a language but not English) and white background.
I hope to use ...
- 21
1
vote
0
answers
108
views
Chicken and egg problem in machine learning [closed]
Recently, I went through an ICLR paper SELF-LABELLING VIA SIMULTANEOUS CLUSTERING
AND REPRESENTATION LEARNING. In the paper, authors discussed simultaneously labeling the images and training a network ...
- 71
1
vote
0
answers
309
views
Any reason for choosing t-SNE over UMAP when visualizing?
According to the UMAP paper:
Our algorithm is competitive with t-SNE for visualization quality and arguably preserves more of the global structure with superior run time performance.
paper
It seems ...
- 153
2
votes
0
answers
76
views
Restricted Boltzmann Machine: W matrix visualization results after training MNIST images and Pseudo-log-likelihood
I am implementing RBM from scratch using Tensorflow and after training my RBM on the MNIST dataset for 200 epochs using Persistent CD with two steps of contrastive divergence, I learn the weights W ...
- 21
0
votes
1
answer
615
views
Do we need to split the data for Unsupervised Anomaly Detection?
I'm struggling with understanding the concept of splitting data for unsupervised anomaly/outlier detection. You can find all approaches here. I found some papers and implementations that didn't split ...
- 407
0
votes
0
answers
55
views
Maximizing a unique trace quadratic form
I am dealing with an unsupervised problem where I have ended up with the following maximization problem:
$\max_{C\in \mathbb{R}^{p\times n}}\sum_{i=0}^{m} tr(CA^ixx^\top A^{i^\top}C^\top) \\ \mathrm{...
- 43