Optimal Number Of Clusters Research Articles

Abstract 17465: Analysis of Intra-Operative Heart Rates Reveals a Novel Sub-Phenotype in Coronary Artery Bypass Grafting Patients

Background: We aim to investigate the use of intra-operative heart rate (HR) trajectories to develop and characterize sub-phenotypes of coronary artery bypass grafting (CABG) patients with distinct risk and outcome profiles. Methods: We used a retrospective cohort including 4194 CABG patients admitted to a large rural healthcare system between 2012 and 2021. Functional data analysis (FDA) was applied to patients’ intra-operative HR trajectories to identify distinct sub-phenotypes. Results: The elbow method suggested that the optimal number of clusters is four. Fig. 1 shows the mean HR trajectory curve (top) and Kaplan-Meier survival curve (bottom) for each of the four identified groups. G1 (Low HR) includes 35.2% of the patients with median age 68 years. Patients in G1 had the lowest ICU after CABG length of stay (LoS) of 76.1 hours and the longest CABG median procedure duration of 4.6 hours. Assignment to G1 was significantly associated with shorter ICU LoS (p= 2.2E-6). G2 (Intermediate HR) includes 35.4% of the patients with median age of 67 years. Patients in G2 had the highest prevalence of urgent admissions, 25.5%. Assignment to G2 was significantly associated with shorter ICU LoS (p=0.016). G3 (Increasing HR) includes 15.6% of the patients with median age 67. Patients in G3 had the shortest median procedure duration of 3.68 hours. Those patients were found to be significantly at risk of prolonged ICU stays (p=0.0002). Finally, G4 (High HR) includes 16.2% with the lowest observed median age of 65 years. Patients in G4 have the highest prevalence of in-hospital mortality (5.2%), 30-day readmission (4.3%), and 1-year mortality (4.8%). Assignment to G4 was significantly associated with the following outcomes: prolonged ICU stays (p=2.62e-9); in-hospital mortality (p=8.03e-5); 30-day readmission (p=0.017); and 1-year mortality (p=0.02). Conclusions: CABG sub-phenotypes based on intra-operative HR trajectories had distinct clinical characteristics and outcomes.

Abstract 14964: Unsupervised Machine Learning of Imaging-Based Sub-Phenotypes of Aortic Stenosis With Differential Risks of Conduction Disturbances After Transcatheter Aortic Valve Replacement

Background: It is currently unclear whether various measurements of the aortic valvular complex routinely performed on pre-operative computed tomography angiograms (CTA) can collectively predict the risk of conduction disturbances after transcatheter aortic valve replacement (TAVR) for aortic stenosis (AS). Here, we aimed to use unsupervised machine learning to analyze a multitude of CTA features and identify distinct patient sub-phenotypes with different risks for TAVR-related conduction disturbances. Methods: The pre-TAVR CTAs of 660 AS patients (330 males, 330 females) were analyzed using TeraRecon to extract 21 features that included aortic valve leaflet calcification loads, dimensions of the aortic root (annulus, sinus of Valsalva, sinotubular junction) and the ascending aorta, coronary ostial heights, and aortic angle. Agglomerative hierarchical clustering was performed separately on male and female datasets using R, with assessment of clusterability based on Hopkins statistics and agglomerative coefficient and the optimal number of clusters based on 30 previously validated indices (NbClust package). Multivariable logistic regression was conducted to assess the dependence of significant conduction disturbances (new left bundle branch block, advanced atrioventricular block) on cluster type. Results: Both male and female datasets were fairly clusterable (Hopkins statistics, 0.70 vs 0.68; agglomerative coefficient, 0.93 vs 0.93), with 3 and 2 as the optimal number of clusters, respectively. Of the 3 male clusters identified, the cluster with intermediate median values of aortic valve leaflet calcification loads and aortic root dimensions (M2) was associated with the greatest risk of conduction disturbances, compared to clusters with low (M1) and high (M3) median values of these variables (OR M2/M1 =9.24, p =0.03; OR M2/M3 =1.33, p =0.40; OR M3/M1 =6.97, p =0.07). The risk of the M1 cluster was not statistically different from those of the female clusters (F1, F2; OR M1/F1 =0.31, p =0.27; OR M1/F2 =0.25, p =0.21), which were not different from each other (OR F1/F2 =0.82, p =0.68). Conclusion: Machine learning of pre-TAVR CTAs can help identify subgroups of patients with similar or varying risks of TAVR-related conduction disturbances.

Abstract 11546: Functional Co-Clustering of Physical Activity Patterns Among Sexual and Gender Minority Adults

Introduction: Physical activity (PA) is a important modifiable risk factor for cardiovascular disease (CVD) and CVD mortality. There is growing evidence that sexual and gender minority (SGM; lesbian/gay, bisexual, transgender) adults are at higher risk of insufficient PA and CVD. Accelerometers allow for collection of granular PA data that can identify individual- and temporal heterogeneity in daily patterns. To date, no studies have investigated accelerometry-based PA patterns among SGM adults. Goal: To characterize daily PA trajectories among SGM adults and identify distinct phenotypes using a time series-based clustering technique. Methods: We recruited an online sample of healthy SGM adults in the United States who completed continuous wrist-worn accelerometry for 30 days to collect step counts. For clustering, we used a functional latent block models (FLBMs), where each 24-hour period was a Fourier-smoothed data curve. FLBMs allow hierarchical time-series data clustering by simultaneously clustering person- and day-level data. The optimal number of clusters was selected using the integrated completed likelihood (ICL) criterion. Results: Forty-two SGM adults with a mean age of 27.0 years (+/- 7.7) provided 1207 person-level days of accelerometry (range=6-31 days/person). A final model with 4 blocks (2 x 2 clusters) provided the best fit (ICL= -70086.4). Two person-level clusters were identified, which were characterized by differences in the amount and distribution of steps throughout the day were identified. Cluster 1 (n=26) had higher overall steps counts with significant morning and evening increases in step counts. Cluster 2 (n=16) had fewer steps that were within a narrower time period. We further identified 2 day-level clusters. Cluster 2 (n=14) had a wider temporal distribution of step counts and a higher variance on weekend days vs. weekdays relative to Cluster 1 (n=17). Conclusions: This is the first study to elucidate daily PA trajectories in SGM people. Using FLBM, we accounted for individual heterogeneity and relations among days. Findings can help identify individuals at increased risk of physical inactivity and subsequent negative health outcomes, providing important knowledge to inform behavioral interventions.

Exploring the heterogeneous effects of riding behaviours and road conditions on delivery rider severities in scooter-style electric bicycle crashes involving vehicles

Delivery riders are more vulnerable than other traffic participants, especially in vehicle-involved delivery crashes. This study aims at identifying the unobserved heterogeneities in different factors, based on 4251 vehicle-scooter-style electric bicycle (SSEB) crashes. First, some potential factors are selected from seven perspectives, and the spatiotemporal characteristics are analysed. Second, a latent class clustering method is proposed to clarify the optimal number of clusters by maximizing the heterogeneities across clusters. Third, partial proportional odds (PPO) models for the whole dataset and sub-datasets are developed to explore the heterogeneities across various clusters. Besides, marginal effects are implemented to quantify the heterogeneities. The results evidence that there are remarkable heterogeneities across different clusters, especially in riding behaviours and road conditions. Several factors only significantly affect particular clusters but not the whole dataset. The PPO models for the sub-datasets perform better in identifying the underlying heterogeneities. The results also highlight the greater roles of riding behaviours and road conditions in delivery SSEB-vehicle crashes. The top five influencing factors are running red light, using cell phones, vehicle type, reverse riding and bike lane (their maximum marginal effects exceeding +35%). The findings could support to mitigate the related crash losses.

CO2 emissions and delivery time of last‐mile drone delivery using trucks

AbstractThe shift in consumer behaviour from in‐person to online shopping has led to an increase in parcel delivery volume and its associated negative impacts, such as CO2 emissions in cities. With the emergence of drone‐delivery technologies, the authors analyzed a joint delivery method using drones and trucks, which is an emerging alternative solution for last‐mile delivery in terms of CO2 emissions and delivery time. The analysis verified the highest possible level of reduction in CO2 emissions for the simultaneous and strategic operation of drones and trucks compared to diesel‐ or electric‐only truck operations. Moreover, this approach leads to reduced delivery times. A sensitivity analysis was performed to optimize the delivery‐drone flight performance in a drone‐and‐truck delivery strategy. It was found that a 1.5‐km drone flight performance was sufficient when considering the reasonable assumptions adopted in this study. Furthermore, based on the analysis results, drone‐and‐truck cooperative delivery strategies may not provide a significant advantage in terms of CO2 emissions compared with alternative transportation modes, such as fuel‐cell trucks with sufficiently low emissions. It has been empirically verified that changes in CO2 emissions are proportional to the number of clusters. However, there is a risk of local optima due to microscopic fluctuations among neighbouring cluster numbers, which occurs during the search for the optimal number of clusters.

Unsupervised Multiview Fuzzy C-Means Clustering Algorithm

The rapid development in information technology makes it easier to collect vast numbers of data through the cloud, internet and other sources of information. Multiview clustering is a significant way for clustering multiview data that may come from multiple ways. The fuzzy c-means (FCM) algorithm for clustering (single-view) datasets was extended to process multiview datasets in the literature, called the multiview FCM (MV-FCM). However, most of the MV-FCM clustering algorithms and their extensions in the literature need prior information about the number of clusters and are also highly influenced by initializations. In this paper, we propose a novel MV-FCM clustering algorithm with an unsupervised learning framework, called the unsupervised MV-FCM (U-MV-FCM), such that it can search an optimal number of clusters during the iteration process of the algorithm without giving the number of clusters a priori. It is also free of initializations and parameter selection. We then use three synthetic and six benchmark datasets to make comparisons between the proposed U-MV-FCM and other existing algorithms and to highlight its practical implications. The experimental results show that our proposed U-MV-FCM algorithm is superior and more useful for clustering multiview datasets.

Clustered photoplethysmogram pulse wave shapes and their associations with clinical data.

Photopletysmography (PPG) is a non-invasive and well known technology that enables the recording of the digital volume pulse (DVP). Although PPG is largely employed in research, several aspects remain unknown. One of these is represented by the lack of information about how many waveform classes best express the variability in shape. In the literature, it is common to classify DVPs into four classes based on the dicrotic notch position. However, when working with real data, labelling waveforms with one of these four classes is no longer straightforward and may be challenging. The correct identification of the DVP shape could enhance the precision and the reliability of the extracted bio markers. In this work we proposed unsupervised machine learning and deep learning approaches to overcome the data labelling limitations. Concretely we performed a K-medoids based clustering that takes as input 1) DVP handcrafted features, 2) similarity matrix computed with the Derivative Dynamic Time Warping and 3) DVP features extracted from a CNN AutoEncoder. All the cited methods have been tested first by imposing four medoids representative of the Dawber classes, and after by automatically searching four clusters. We then searched the optimal number of clusters for each method using silhouette score, the prediction strength and inertia. To validate the proposed approaches we analyse the dissimilarities in the clinical data related to obtained clusters.

Homogenous Climatic Regions for Targeting Green Water Management Technologies in the Abbay Basin, Ethiopia

Spatiotemporal climate variability is a leading environmental constraint to the rain-fed agricultural productivity and food security of communities in the Abbay basin and elsewhere in Ethiopia. The previous one-size-fits-all approach to soil and water management technology targeting did not effectively address climate-induced risks to rain-fed agriculture. This study, therefore, delineates homogenous climatic regions and identifies climate-induced risks to rain-fed agriculture that are important to guide decisions and the selection of site-specific technologies for green water management in the Abbay basin. The k-means spatial clustering method was employed to identify homogenous climatic regions in the study area, while the Elbow method was used to determine an optimal number of climate clusters. The k-means clustering used the Enhancing National Climate Services (ENACTS) daily rainfall, minimum and maximum temperatures, and other derived climate variables that include daily rainfall amount, length of growing period (LGP), rainfall onset and cessation dates, rainfall intensity, temperature, potential evapotranspiration (PET), soil moisture, and AsterDEM to define climate regions. Accordingly, 12 climate clusters or regions were identified and mapped for the basin. Clustering a given geographic region into homogenous climate classes is useful to accurately identify and target locally relevant green water management technologies to effectively address local-scale climate-induced risks. This study also provided a methodological framework that can be used in the other river basins of Ethiopia and, indeed, elsewhere.

Implementasi Clustering K-Medoids dalam Pengelompokan Kabupaten di Provinsi Aceh Berdasarkan Faktor yang Mempengaruhi Kemiskinan

<p><em>The economy is one of the parameters to see how the development of a country. Ending poverty anywhere and in any form is goal 01 of the Sustainable Development Goals (SDGs) program. Until now, poverty has become one of the main problems in Indonesia, so poverty must be a concern of the government. Based on data from the Central Statistics Agency (BPS) shows that as of September 2020 the percentage of poor people in Aceh Province is still the highest on the island of Sumatra, which is 15.43%. The purpose of this study is to classify districts based on factors that affect poverty in Aceh Province. The method used in this study is the K-Medoids Cluster Analysis algorithm. The optimal number of clusters is 2 clusters with cluster 1 consisting of 11 districts and cluster 2 consisting of 12 districts. Cluster 1 has a higher percentage of poor population and poverty depth index than cluster 2, while cluster 2 has higher Gini Ratio, AHH, and RLS values than cluster 1.</em></p><p><em></em><strong><em>Keywords : </em></strong><em>Clusters, Economy, Poverty, SDGs</em></p>

KClusterHub: An AutoML-Driven Tool for Effortless Partition-Based Clustering over Varied Data Types

Partition-based clustering is widely applied over diverse domains. Researchers and practitioners from various scientific disciplines engage with partition-based algorithms relying on specialized software or programming libraries. Addressing the need to bridge the knowledge gap associated with these tools, this paper introduces kClusterHub, an AutoML-driven web tool that simplifies the execution of partition-based clustering over numerical, categorical and mixed data types, while facilitating the identification of the optimal number of clusters, using the elbow method. Through automatic feature analysis, kClusterHub selects the most appropriate algorithm from the trio of k-means, k-modes, and k-prototypes. By empowering users to seamlessly upload datasets and select features, kClusterHub selects the algorithm, provides the elbow graph, recommends the optimal number of clusters, executes clustering, and presents the cluster assignment, through tabular representations and exploratory plots. Therefore, kClusterHub reduces the need for specialized software and programming skills, making clustering more accessible to non-experts. For further enhancing its utility, kClusterHub integrates a REST API to support the programmatic execution of cluster analysis. The paper concludes with an evaluation of kClusterHub’s usability via the System Usability Scale and CPU performance experiments. The results emerge that kClusterHub is a streamlined, efficient and user-friendly AutoML-inspired tool for cluster analysis.

Market segmentation: Machine Learning in Marketing in the Context of COVID-19

The COVID-19 health crisis has led to unprecedented changes in consumer behavior, as consumers now purchase differently and use different means. Consumers are checking and judging products via electronic devices, shaping trends in consumer segments. This research study aimed to use the clustering model with Machine Learning resources in the analysis of clusters as a resource for consumer segmentation, a major component in business marketing management. A 6-question questionnaire was administered to 506 people ranging from 18 to 65 years old to gauge their opinions about going shopping. A dataset was organized using the data collected and processed using RapidMiner Studio 9.10 software. The optimal number of clusters and their components were obtained from the performance indicator provided by Machine Learning.

Towards a precision medicine Solution for optimal pediatric Laparoscopy: An exploratory data analysis for features Selections

Over the past years, healthcare methodologies were commonly applied to all patients affected by the same pathology as an overall treatment. However, recent scientific works were mainly aimed at personalized care, according to the patient’s clinical conditions and morphological characteristics. Pediatric surgery was also involved in this process, although laparoscopy has been set aside so far. In this regard, the physical characteristic that correlates intra-abdominal pressure and intra-abdominal volume (i.e., abdominal com-pliance) was always studied as a unique curve. This work aims to propose an alternative approach to treat laparoscopy surgery and abdominal compli-ance for pediatric individuals, which is based on personalized medicine. We perform an exploratory data analysis considering a real dataset, in order to study the data model (i.e., descriptive univariate and multivariate analysis) and the predictors (i.e., variance thresholding, Shapiro-Wilk test, Pearson’s correlation coefficient) that will allow physicians to build clusters of pediatric individuals. Data analysis proves that there are strong correlations among features referring to thorax circumference and xifo-bisialic lenght with the sum of thorax and pelvis volumes. This leads to identify the optimal number of clusters through the Elbow method.

Combining contour-based and region-based in image segmentation

Background: This paper presents an optimized clustering approach applied to image segmentation. Accurate image segmentation impacts many fields like medical, machine vision, object detection. Applications involve tumor detection, face detection and recognition, and video surveillance. Methods: The developed approach is based on obtaining an optimum number of clusters and regions of an image. We combined Region-based and contour-based approaches. Initial rough regions are obtained using edge detection. We have used Gabor wavelets for texture classification and spatial resolutions. Color frequencies are also used to determine the number of clusters of the Fuzzy c-means (FCM) algorithm which gave an optimum number of clusters or regions. Results: We have compared our approach with other similar wavelet and clustering techniques. Our algorithm gave better values for segmentation metrics like SNR, PSNR, and MCC. Conclusions: Optimizing the number of clusters or regions has a significant effect on the performance of the image segmentation techniques. This will result in better detection and localization of the segmentation-based application.

Random clustering and dynamic recognition-based operation strategy for energy storage system in industrial park

The high volatility and intermittency of power load pose significant challenges to achieving optimal operation of energy storage system (ESS), which ultimately affects the economic benefits of industrial parks. To address this issue, this paper proposes a random clustering and dynamic recognition-based operation strategy for ESS in industrial parks. Firstly, we propose an expected cost minimization-driven random clustering method to determine the optimal cluster number of load curves. We then develop optimal day-ahead self-operating strategies of ESS corresponding to each of the clustered load curves. The uncertainty of power load is described by linearized risk assessment indexes. Secondly, we present a dynamic recognition technology that consists of the characteristics of value range, slope similarity, power size similarity, and curve step similarity to recognize the cluster to which the intra-day load curve belongs. Finally, we select the optimal intra-day charging/discharging strategy of ESS according to the recognized cluster as the operation strategy of the industrial park. Numerical experiments on a real-life industrial park with photovoltaic and ESS validate that the proposed strategy can reduce operation costs as well as promote photovoltaic local accommodation.

Research on the Application of Civic Education into the Brain and Heart Based on Physiological Data Recognition Algorithm

Abstract This paper designs a Civic Education model based on a physiological data identification algorithm, highlighting the training enhancement of physiological data mining Civic Education for students. In the design process, the clustering similarity is calculated based on the attribute distance, and the optimal number of clusters is selected using the contour coefficient. On the basis of determining the optimal number of clusters, the maximum number of iterations is performed, and the output at termination is used to divide the thinking and teaching ability and ability assessment index association distinction. The minimum confidence threshold association rule is developed through the candidate item support, and weighted random selection is performed for each dimension. To verify the feasibility of this educational model, its teaching effect is analyzed. The experimental results show that the accuracy of the output results is high, and the average time of Civic Education of the philosophy of dealing with the world is about 0.65 seconds. The highest psychological recall rate of Civic Education reached 0.59. It shows that physiological data mining can guide students to build a Civic Education thinking system so that the course and Civic Education can form a synergistic effect, thus forming a horizontal and vertical synergistic education situation.

Medoid Silhouette clustering with automatic cluster number selection

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, provide two fast versions for the direct optimization, and discuss the use to choose the optimal number of clusters. 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(k2). In experiments on real data with 30000 samples and k= 100, we observed a 10464× speedup compared to the original PAMMEDSIL algorithm. Additionally, we provide a variant to choose the optimal number of clusters directly.

A Primer on Clustering of Forest Management Units for Reliable Design-Based Direct Estimates and Model-Based Small Area Estimation

This study employs clustering analysis to group forest management units using auxiliary, satellite imagery-derived height metrics and past wall-to-wall tree census data from a natural, uneven-aged forest. Initially, we conducted an exhaustive exploration to determine the optimal number of clusters k, considering a wide range of clustering schemes, indices, and two specific k ranges. The optimal k is influenced by various factors, including the minimum k considered, the selected clustering algorithm, the clustering indices used, and the auxiliary variables. Specifically, the minimum k, the Euclidean distance metric, and the clustering index were instrumental in determining the optimal cluster numbers, with algorithms exerting minimal influence. Unlike traditional validation indices, we assessed the performance of these optimally defined clusters based on direct estimates and additional criteria. Subsequently, our research introduces a twofold methodology for Small Area Estimation (SAE). The first approach focuses on aggregating forest management units at the cluster level to increase the sample size, thereby yielding reliable design-based direct estimates for key forest attributes, including growing stock volume, basal area, tree density, and mean tree height. The second approach prepares area-level data for the future application of model-based estimators, contingent on establishing a strong correlation between target and auxiliary variables. Our methodology has the potential to enhance forest inventory practices across a wide range of forests where area-level auxiliary covariates are available.

Model-based clustering of multipath propagation in powerline communication channels

Powerline communication (PLC) channels are known to exhibit multipath propagation behaviour. The authors present a model-based framework to address the challenge of clustering multipath propagation components (MPCs) in PLC channels for indoor low-voltage (LV) environments. The framework employs a range of finite-mixture models (FMMs), including the gamma mixture model, the inverse gamma mixture model, the Gaussian mixture model, the inverse Gaussian mixture model, the Nakagami mixture model, the inverse Nakagami mixture model (INMM) and the Rayleigh mixture model, to identify clusters of MPCs. A measurement campaign of an unknown indoor LV PLC channel is conducted to obtain a channel response. From the channel response, the delay and magnitude parameters of the MPCs are extracted using the space-alternating generalised expectation maximisation algorithm adopted only for these parameters. A maximum likelihood approach and the expectation–maximisation algorithm are employed to fit the FMMs to the MPC delay-magnitude dataset to cluster MPCs in the delay domain. The results of the model-fitting process are then evaluated using the corrected Akaike information criterion (AICc), which enables a fair comparison of the candidate models over the feasible and finite range of clusters. A novel algorithm is introduced for estimating the feasible and finite range of clusters using the extracted delay and magnitude MPC parameters. The AICc’s ranking results show that the INMM model provides the best fit. Davies–Bouldin (DB) and Calinski–Harabasz (CH) indexes are used to compare the model-based clustering approach to the conventional distance-based clustering methods. Validation results show that CH and DB indexes closely agree in the optimal number of MPC clusters for model-based clustering, which corresponds to the most within-cluster compactness of MPCs and to the most between-cluster separation in the delay domain.

CLUSTERIZATION OF REGION IN SOUTH SUMATERA BASED ON COVID-19 CASE DATA

Based on Covid-19 case data as of July 2022, South Sumatra Province has the 15th highest rank out of 34 provinces in Indonesia, with confirmed cases totalling 82,407. This showed that the spread of Covid-19 in South Sumatra was still high. This study aimed to determine the cluster of regions in South Sumatra based on Covid-19 case data. Clustering regions used agglomerative hierarchical method. The process began with standardizing the data, calculating the similarity distance between objects, determining the optimal number of clusters using the Silhouette method, and the last was clustering analysis. This study found that the optimal number of clusters consisted of two clusters. The clustering process starts with objects 2 and objects 4 because these two objects have the closest similarity distance. In conclusion, objects with the closest similarity distance (in one cluster) have the same data movement (fluctuation).

Work Motivation Profiles of the Millennial Generation

Purpose: This study aimed to determine characteristic profiles of the Millennial generation based on their sociodemographic features and motivational preferences regarding work. It contributes to the literature on Millennial motivation and provides insights for organizations seeking to better understand and manage said generation. Design/Methodology: The study was conducted on a sample of 197 questionnaire responses from individuals in the Millennial generation who had work experience. The sampling was non-probabilistic and did not consider aspects related to socioeconomic or education levels to broaden the coverage of the study. The data were collected through an online survey in Guadalajara, Jalisco, Mexico. Said data were examined using an analytical procedure—which involves a clustering algorithm to determine the optimal number of clusters—and logistic regression analysis—to identify significant variables that can explain the behavior of each group. Findings: Two distinct motivational profiles were found among Millennials: (1) a group motivated by achievement and power and (2) another one inspired by affiliation and supervision group. It was also found that these two profiles are related to certain sociodemographic features, such as age and main breadwinner. Conclusions: Understanding the motivational profiles of Millennials can help organizations better tailor their management practices and work environments to meet the needs of this generation. Likewise, organizations may need to provide several kinds of incentives and rewards to motivate different groups of Millennials. Future research in this area could explore the relationship between these motivational profiles and other outcomes, such as job satisfaction and turnover. Originality: This study contributes to the literature on Millennial motivation by introducing a quantitative methodology to identify different motivational profiles and explore their relationship with sociodemographic features. The use of a clustering algorithm and regression analysis also contributes to the methodological approaches employed in this area of research. Focused on the Mexican context, this paper also provides insights into the unique cultural and economic factors that may influence Millennial motivation in this region.