The evaluation of clustering results is difficult, highly dependent on the evaluated data set and the perspective of the beholder. There are many different clustering quality measures, which try to provide a general measure to validate clustering results. A very popular measure is the Silhouette. We discuss the efficient medoid-based variant of the Silhouette, perform a theoretical analysis of its properties, and provide two fast versions for the direct optimization. We combine ideas from the original Silhouette with the well-known PAM algorithm and its latest improvements FasterPAM. One of the versions guarantees equal results to the original variant and provides a run speedup of $O(k^2)$. In experiments on real data with 30000 samples and $k$=100, we observed a 10464$\times$ speedup compared to the original PAMMEDSIL algorithm.

Due to the rapid development of non-face-to-face services due to the corona virus, commerce through the Internet, such as sales and reservations, is increasing very rapidly. Consumers also post reviews, suggestions, or judgments about goods or services on the website. The review data directly used by consumers provides positive feedback and nice impact to consumers, such as creating business value. Therefore, analysing review data is very important from a marketing point of view. Our research suggests a new way to find factors for customer satisfaction through review data. We applied a method to find factors for customer satisfaction by mixing and using the data mining technique, which is a big data analysis method, and the natural language processing technique, which is a language processing method, in our research. Unlike many studies on customer satisfaction that have been conducted in the past, our research has a novelty of the thesis by using various techniques. And as a result of the analysis, the results of our experiments were very accurate.

In medicine, curated image datasets often employ discrete labels to describe what is known to be a continuous spectrum of healthy to pathological conditions, such as e.g. the Alzheimer's Disease Continuum or other areas where the image plays a pivotal point in diagnosis. We propose an architecture for image stratification based on a conditional variational autoencoder. Our framework, VAESim, leverages a continuous latent space to represent the continuum of disorders and finds clusters during training, which can then be used for image/patient stratification. The core of the method learns a set of prototypical vectors, each associated with a cluster. First, we perform a soft assignment of each data sample to the clusters. Then, we reconstruct the sample based on a similarity measure between the sample embedding and the prototypical vectors of the clusters. To update the prototypical embeddings, we use an exponential moving average of the most similar representations between actual prototypes and samples in the batch size. We test our approach on the MNIST-handwritten digit dataset and on a medical benchmark dataset called PneumoniaMNIST. We demonstrate that our method outperforms baselines in terms of kNN accuracy measured on a classification task against a standard VAE (up to 15% improvement in performance) in both datasets, and also performs at par with classification models trained in a fully supervised way. We also demonstrate how our model outperforms current, end-to-end models for unsupervised stratification.

A rapid transformation of current electric power and natural gas (NG) infrastructure is imperative to meet the mid-century goal of CO2 emissions reduction requires. This necessitates a long-term planning of the joint power-NG system under representative demand and supply patterns, operational constraints, and policy considerations. Our work is motivated by the computational and practical challenges associated with solving the generation and transmission expansion problem (GTEP) for joint planning of power-NG systems. Specifically, we focus on efficiently extracting a set of representative days from power and NG data in respective networks and using this set to reduce the computational burden required to solve the GTEP. We propose a Graph Autoencoder for Multiple time resolution Energy Systems (GAMES) to capture the spatio-temporal demand patterns in interdependent networks and account for differences in the temporal resolution of available data. The resulting embeddings are used in a clustering algorithm to select representative days. We evaluate the effectiveness of our approach in solving a GTEP formulation calibrated for the joint power-NG system in New England. This formulation accounts for the physical interdependencies between power and NG systems, including the joint emissions constraint. Our results show that the set of representative days obtained from GAMES not only allows us to tractably solve the GTEP formulation, but also achieves a lower cost of implementing the joint planning decisions.

In the new era of digitalisation, there are emerging challenges relating to the new technology risks while fraud risk increases, adopting new and innovative methods in line with technological progress. In this environment, audit should develop data analytics skills that are beyond traditional risk monitoring and fraud detection tools to meet stakeholders' evolving expectations. Clustering is an unsupervised Machine Learning (ML) algorithm (i.e. an algorithm that learns and improves from experience, without input from users) that looks for patterns in data by dividing it into clusters. These clusters are created such that the points are homogenous within the cluster and heterogenous across clusters. Clustering is commonly used in market segmentation and several areas of marketing analytics as well as in fraud detection.

This paper introduces an algorithm for the detection of change-points and the identification of the corresponding subsequences in transient multivariate time-series data (MTSD). The analysis of such data has become more and more important due to the increase of availability in many industrial fields. Labeling, sorting or filtering highly transient measurement data for training condition based maintenance (CbM) models is cumbersome and error-prone. For some applications it can be sufficient to filter measurements by simple thresholds or finding change-points based on changes in mean value and variation. But a robust diagnosis of a component within a component group for example, which has a complex non-linear correlation between multiple sensor values, a simple approach would not be feasible. No meaningful and coherent measurement data which could be used for training a CbM model would emerge. Therefore, we introduce an algorithm which uses a recurrent neural network (RNN) based Autoencoder (AE) which is iteratively trained on incoming data. The scoring function uses the reconstruction error and latent space information. A model of the identified subsequence is saved and used for recognition of repeating subsequences as well as fast offline clustering. For evaluation, we propose a new similarity measure based on the curvature for a more intuitive time-series subsequence clustering metric. A comparison with seven other state-of-the-art algorithms and eight datasets shows the capability and the increased performance of our algorithm to cluster MTSD online and offline in conjunction with mechatronic systems.

Regionalization is the act of breaking a dataset into contiguous homogeneous regions that are heterogeneous from each other. Many different algorithms exist for performing regionalization; however, using these algorithms on large real world data sets have only become feasible in terms of compute power in recent years. Very few studies have been done comparing different regionalization methods, and those that do lack analysis in memory, scalability, geographic metrics, and large-scale real-world applications. This study compares state-of-the-art regionalization methods, namely, Agglomerative Clustering, SKATER, REDCAP, AZP, and Max-P-Regions using real world social determinant of health (SDOH) data. The scale of real world SDOH data, up to 1 million data points in this study, not only compares the algorithms over different data sets but provides a stress test for each individual regionalization algorithm, most of which have never been run on such scales previously. We use several new geographic metrics to compare algorithms as well as perform a comparative memory analysis. The prevailing regionalization method is then compared with unconstrained K-Means clustering on their ability to separate real health data in Virginia and Washington DC.

The mutual incompatibility of distinct spectroscopic systems is among the most limiting factors in Laser-Induced Breakdown Spectroscopy (LIBS). The cost related to setting up a new LIBS system is increased, as its extensive calibration is required. Solving the problem would enable inter-laboratory reference measurements and shared spectral libraries, which are fundamental for other spectroscopic techniques. In this work, we study a simplified version of this challenge where LIBS systems differ only in used spectrometers and collection optics but share all other parts of the apparatus, and collect spectra simultaneously from the same plasma plume. Extensive datasets measured as hyperspectral images of heterogeneous specimens are used to train machine learning models that can transfer spectra between systems. The transfer is realized by a pipeline that consists of a variational autoencoder (VAE) and a fully-connected artificial neural network (ANN). In the first step, we obtain a latent representation of the spectra which were measured on the Primary system (by using the VAE). In the second step, we map spectra from the Secondary system to corresponding locations in the latent space (by the ANN). Finally, Secondary system spectra are reconstructed from the latent space to the space of the Primary system. The transfer is evaluated by several figures of merit (Euclidean and cosine distances, both spatially resolved; k-means clustering of transferred spectra). The methodology is compared to several baseline approaches.

Machine Learning problems deal with a great deal of data and depend heavily on the algorithms that are used to train the model. There are various approaches and algorithms to train a machine learning model based on the problem at hand. Supervised and unsupervised learning are the two most prominent of these approaches. An important real-life problem of marketing a product or service to a specific target audience can be easily resolved with the help of a form of unsupervised learning known as Clustering. This article will explain clustering algorithms along with real-life problems and examples.

A keyword search on constrained clustering on Web-of-Science returned just under 3,000 documents. We ran automatic analyses of those, and compiled our own bibliography of 183 papers which we analysed in more detail based on their topic and experimental study, if any. This paper presents general trends of the area and its sub-topics by Pareto analysis, using citation count and year of publication. We list available software and analyse the experimental sections of our reference collection. We found a notable lack of large comparison experiments. Among the topics we reviewed, applications studies were most abundant recently, alongside deep learning, active learning and ensemble learning.