Questions tagged [unsupervised-learning]

Finding hidden (statistical) structure in unlabelled data, including clustering and feature extraction for dimensionality reduction.

0
votes
1answer
22 views

Searching for the correct approach

I'm writing my Bachelor Thesis within the field of neural networks and I need some preparation of the data I'm using. Do you have an idea how I am able to identify the four levels in this graph via ...
2
votes
2answers
210 views

Unsupervised Learning: Train Test division

I have one conceptual question. In Unsupervised Learning, when I have no labels. The anomaly detection model (Isolation forests, Autoencoders, Distance-based methods etc.), it should fit on a ...
1
vote
1answer
234 views

When does my autoencoder start to overfit?

I am working on anomaly detection using an autoencoder neural network with $1$ hidden layer. This is an unsupervised setting, as I do not have previous examples of anomalies. The input data has ...
2
votes
0answers
33 views

Understanding short animation about Dirichlet Process Mixture Model

On the wikipedia page of Dirichlet Process, there is the following video. I don't get the point of the video. My first impression was that the video was showing the fitting of one-dimensional data ...
3
votes
2answers
52 views

Confusion in modelling finite mixture model

From the book "Machine Learning a probabilistic Perspective", I'm reading about finite/infinite mixture models. Particularly at paragraph 25.2.1 it's stated: The usual representation (of a finite ...
2
votes
1answer
112 views

Custom Loss Function - Inducing sparsity

From the comments, I realized that my question wasn't clear enough, so I'll start with a short background. I am trying to construct an attention model that performs classification based on just a ...
2
votes
0answers
41 views

Performing clustering without a distance matrix

I have n vectors and a matrix of similarity scores between them (e.g. vector 1 score of similarity with vector 4 is 1.3, and with vector 7 is 2.3). This matrix is ...
2
votes
0answers
243 views

Clustering as a method to find and label classes for supervised learning

I'm working on a text classification project. We have around 300k documents (small, 1~2 phrases) and we don't know the set of labels or how many labels there are. The recommended approach to me was ...
1
vote
1answer
74 views

Latent variable in Gaussian Mixture Model

Whenever I look up material pertaining to Gaussian Mixture Models, it always mentions latent variable $z$, where $z \in \{1, ..., K\}$ and is one-hot encoded. I completely understand the objective of ...
0
votes
3answers
239 views

Why are decision trees considered supervised learning?

It seems to work similar to clustering algorithms, where data does not have to be labeled, and the algorithm creates it's own labels/groups based on feature similarities...
1
vote
0answers
59 views

clustering objects in point cloud

I am currently working on point cloud data analysis, trying to label objects which are not ground or vegetation e.t.c. So far I tried many clustering algorithms, with moderate success. In my best ...
0
votes
0answers
22 views

DBSCAN loops one or several times a data point?

I am trying to construct a model data to facilitate the clustering algortihms execution in terms of searching for data point in the dataset. This model is a set of connections between points such that ...
3
votes
2answers
92 views

External validation of clustering requires labels, but why cluster at all if you have labels?

There are two types of validation in clustering, using: Internal indexes: Used to measure the goodness of a clustering structure without respect to external information (e.g., sum of squared errors)...
1
vote
1answer
28 views

How can I tell if an Autoencoder is encoding my data properly?

Autoencoders can be classified as a method of unsupervised learning, and unsupervised learning often comes with a problem where it's hard to tell if the model is working properly. However, some ...
1
vote
0answers
118 views

Clustering Noisy Data [closed]

There are various ways to cluster data. Some require the data first to be scaled to have a mean of $0$ and standard deviation of $1$. However, others do not mention if the data should be scaled at all....
1
vote
1answer
152 views

Is clustering subsequences of time-series still meaningless with unsupervised learning?

In the paper "Clustering of Time Series Subsequences Is Meaningless" Keoh et al. claim that breaking a time-series into chunks (sometimes called lags) of fixed-size using the rolling window method ...
1
vote
1answer
163 views

Clustering data that includes a categorical variable with many different levels

I'm looking to cluster data on apartments. I have the following variables for each apartment: Latitude Longitude Price Number of bathrooms Number of bedrooms Amenities (washer, gym, etc.) The ...
1
vote
0answers
23 views

Performance of Hierarchical Temporal Memory on unsupervised online anomaly detection problems

I'm looking for an algorithm to detect anomalies in streaming data (server metrics). The detection needs to be near-real time and unsupervised (labeled data will never be available, unfortunately, and ...
2
votes
0answers
16 views

What are the simple methods to do an unsupervised cluster to stock return time series?

I am a student in finance and I am working on my thesis project. I am interested in doing a clustering to stock time series. I first read the paper 'Time-series clustering – A decade review' from ...
2
votes
3answers
89 views

How to find strongly connected subgraphs in a graph? [duplicate]

I have a simple, undirected graph where I'd like to detect "natural" subgraphs where vertices are connected intensively internally but sparsely externally. The problem is that I have no exact hint ...
0
votes
0answers
16 views

What are some good probabilistic modeling approaches for co-occurance of words in large corpus of short documents?

I'd be very grateful for some advice in finding a modeling approach and toolset for a NLP problem I have. I need to learn a joint distribution of tokens in a corpus of docs. I have on the order of ...
0
votes
0answers
158 views

Action space for Reinforcement Learning implementation

I am confused about how to set action space for my application for which I like to use Reinforcement Learning to select the best instance. I have two groups of instances. Group_a and Group_b, ...
2
votes
0answers
124 views

what's the difference between semi-supervised learning and partially supervised learning? [closed]

Isn't every semi-supervised problem also a partially supervised learning problem and vice versa?
-1
votes
1answer
15 views

Which clustering technique to use for 1D data with 2 Gaussian distributions without training data?

I have 1D data which is likely to be 2 Gaussian distributions overlapped. I do not have training data but I know one should be slightly high contributed from few points and another is very broad with ...
-1
votes
1answer
29 views

K-means in R: complete case analysis followed by nearest-neighbor assignment for partial data

I have a dataset of 3K observations with only 162 being a complete case. I have read here that it is possible to run knn on the complete cases and then conduct a nearest neighbour assignment for ...
1
vote
1answer
110 views

Is a chi-square test for independence appropriate on a contingency table where one category is the unsupervised learning cluster?

I have a data set that has been partitioned into four clusters by executing a clustering algorithm that used principal components from a principal component analysis (PCA). I then make a contingency ...
1
vote
0answers
23 views

When applying PCA to a dataset consisting of regression coefficients, should one use PCA on correlation or on covariance?

This is a follow-up question from the post: PCA on correlation or covariance? The accepted answer quotes: You tend to use the covariance matrix when the variable scales are similar and the ...
0
votes
1answer
130 views

How to get meaningful results from Softmax activation in Deep Unsupervised Clustering Network

I found this interesting paper regarding deep unsupervised clustering and am looking to mimic some of the things done, however there is one thing that is not clear to me. In the paper, they use a ...
1
vote
0answers
55 views

What exactly is semisupervised learning?

I have come across two descriptions of what semisupervised learning is, where one would have a small set $\mathcal{L}$ of labeled data and a larger set $\mathcal{U}$ of unlabeled data. The first ...
0
votes
1answer
39 views

Cross Validation in application of clustering on a collection of similarly behaving time series

I'm trying to understand how and at which point can one apply Cross Validation for time series data. If i'm not wrong CV increases generalisation so that our model has less bias in case the data is ...
1
vote
2answers
45 views

Unsupervised learning with missing features

Assume I have a set of N vectors with M features each. If I want to create a latent space to project these vectors into, there are a variety of techniques available to me, such as Principle Component ...
3
votes
1answer
3k views

K-nearest neighbor supervised or unsupervised machine learning?

I've read in several papers that K-nearest neighbor can be supervised or unsupervised learning. Is Knn always unsupervised when one use it for clustering and supervised when one used it for ...
-1
votes
1answer
81 views

clustering VS supervised classification, in the case of very small database

I'm trying to classify/cluster subjects according to 4 features in two classes: healthy and sick. Two things to know: I know the labels/classes of each subject + I only have 40 subjects (in total: ...
1
vote
0answers
128 views

Unsupervised Text Clustering Project: How to get started? [closed]

I work for a manufacturing company where robust databases and data integrity have not always been a priority. I have a very messy and finite list of 13,102 tool descriptions. I need to find out how ...
0
votes
1answer
211 views

unsupervised classification VS supervised classification when data labels are known

Can someone give me some scenario where it's better to use clustering (unsupervised classification) than supervised classification such as SVM ? I mean in a case where you know the data labels/classes....
14
votes
3answers
494 views

What *is* an Artificial Neural Network?

As we delve into Neural Networks literature, we get to identify other methods with neuromorphic topologies ("Neural-Network"-like architectures). And I'm not talking about the Universal Approximation ...
1
vote
1answer
98 views

Is it possible to select features from completely unlabeled data?

I have seen many examples of using semi-supervised learning to reduce the the number of features in a data set, but I am wondering if it is possible to somehow reduce features with purely unlabeled ...
1
vote
0answers
25 views

how many data points for effective LDA (clustering)?

Is there an advisable number of data points recommended to be able to effectively use LDA (Linear discriminant analysis)? I know this is a very open-ended question: but how many is enough?
2
votes
1answer
108 views

“Reverse” clustering?

There is a lot of content about how to cluster, say, customers (k-means, EM clustering, etc.). However, is there a way to reverse cluster customers? Meaning, let's say I have 20 customers, and want ...
0
votes
1answer
373 views

Training and testing an autoencoder on very sparsely populated data

I am exploring the possibility of using a deep autoencoder neural net to build a recommender system. I am firstly testing the model's performance on the traditionally used benchmark of the Movielens ...
0
votes
0answers
97 views

Factor Scores and Clustering

I used self organizing map to cluster factor scores. I used 5 variables for clustering and at the end, I have 4 clusters. Some factor score means are negative. How am I supposed to interpret these ...
0
votes
1answer
157 views

Difference between Hartigan & Wong Algo to Lloyd's algorithm in K-means clustering

In the iterations of Hartigan and Wong Algo of K-Means clustering, If the centroid is updated in the last step, for each data point included, the within- cluster sum of squares for each data point if ...
2
votes
0answers
638 views

Unsupervised anomaly detection - metric for tuning Isolation Forest parameters

I have a project, in which, one of the stages is to find and label anomalous data points, that are likely to be outliers. As a first step, I am using Isolation Forest algorithm, which, after plotting ...
0
votes
0answers
101 views

agglomerative clustering of undirected weighted graph

Let's say I have a simple graph of undirected, weighted edges. I want to agglomerate nodes one-at-a-time by combining the two nodes which have the highest weighted edge, recalculating remaining edge ...
0
votes
1answer
74 views

What is the best approach of clustering if you know number of elements in each cluster?

I want to predict sporting event with knockout system tournament. Data have some features attached to each athlete. I know that there are determined number of results such as: ...
0
votes
0answers
21 views

Get rid of the disrupting points

Here on the following picture, we can identify two clouds. We can identify a lot of points that are out of the two clouds and can disrupt a bit the the training and test accuracy (%). Is there an ...
2
votes
2answers
521 views

Kmeans - Does removing outliers on a dimension affect other dimensions clustering prediction?

I have a set with several features that I wish to cluster using Kmeans. If I remove a point that is an outlier on one dimenssion but not in the others will it affect the result? Outliers were found ...
0
votes
1answer
26 views

Should feature scaling be used while using unsupervised algorithms?

I have read many articles and resources about using feature scaling and when to use it, in particular two answers on this website as well- When should I apply feature scaling for my data? What ...
0
votes
0answers
46 views

feature score(or importance) in each cluster

I want to categorize my clusters with the features present in it. I have 4 clusters and want to know to what extent a feature is important to its cluster. I need a valid scoring metric that tells the ...
1
vote
0answers
28 views

How to evaluate the performance of a predictor on an unlabelled dataset? What is an appropriate test set size and how to sample it?

I am working on a project with goal to deduplicate a customer database. We don't have any annotations (no training/test set with the ground truth values). We implemented multiple unsupervised ...