Buy Ecstasy online in Santa Marta

__________________________

📍 Verified store!

📍 Guarantees! Quality! Reviews!

__________________________

▼▼ ▼▼ ▼▼ ▼▼ ▼▼ ▼▼ ▼▼

▲▲ ▲▲ ▲▲ ▲▲ ▲▲ ▲▲ ▲▲

Buy Ecstasy online in Santa Marta

This is a HealthLine Online content page created by Healthwise. HealthLine Online helps you make better decisions about your health. MDMA is a stimulant. It also has mild hallucinogenic effects. It's often called ecstasy or Molly. MDMA is most often taken as a pill. The pills often have a logo, such as a cartoon character, stamped on them. MDMA also comes as a capsule or powder, which can be snorted, or as a liquid, which can be injected into a vein. Like other stimulants, MDMA can raise a person's heart rate and blood pressure. It increases alertness and may create a feeling of euphoria. It may cause mild hallucinations or seeing, hearing, and feeling things that seem real but aren't. Other effects may include feelings of peacefulness, acceptance, and empathy. MDMA can cause unpleasant side effects, such as:. MDMA can cause confusion, depression, sleep problems, and severe anxiety that may last weeks after taking the drug. Over time, using MDMA can lead to thinking and memory problems. In high doses, MDMA can cause a sharp rise in body temperature. This can lead to serious or even deadly problems such as liver, kidney, or heart failure. A person who doesn't drink enough fluids can become severely dehydrated. The effects of MDMA can be more harmful when it is used with alcohol. MDMA usually does not last in a person's system longer than 12 to 16 hours. And many general drug screening tests do not detect it unless it is specifically targeted. Signs that a person may be using MDMA include:. Author: Ignite Healthwise, LLC Staff Clinical Review Board All Healthwise education is reviewed by a team that includes physicians, nurses, advanced practitioners, registered dieticians, and other healthcare professionals. Clinical Review Board All Healthwise education is reviewed by a team that includes physicians, nurses, advanced practitioners, registered dieticians, and other healthcare professionals. This information does not replace the advice of a doctor. Ignite Healthwise, LLC, disclaims any warranty or liability for your use of this information. Your use of this information means that you agree to the Terms of Use. Learn how we develop our content. If you have questions about your health, dial on your phone or visit HealthLine Online. Top of the page. Overview MDMA is a stimulant. MDMA can cause unpleasant side effects, such as: Muscle tension and jaw-clenching. Blurred vision. Chills or sweating. Skin rash similar to acne. Having a powdered substance or pills stamped with cartoon or other characters. Personality changes. Credits Current as of: November 15, Current as of: November 15, Help Information Emergency. If you believe you have an emergency, dial

How Easy Is It To Get Cocaine in Colombia?

Buy Ecstasy online in Santa Marta

Official websites use. Share sensitive information only on official, secure websites. Competing Interests: The authors have declared that no competing interests exist. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. The number of health-related incidents caused using illegal and legal psychoactive substances PAS has dramatically increased over two decades worldwide. In Colombia, the use of illicit substances has increased up to It is well-known that identifying drug consumption patterns in the general population is essential in reducing overall drug consumption. To enhance our understanding of mental health issues related to PAS and assist in the development of national policies, here we present a novel Deep Neural Network-based Clustering-oriented Embedding Algorithm that incorporates an autoencoder and spatial techniques. The primary goal of our model is to identify general and spatial patterns of drug consumption and abuse, while also extracting relevant features from the input data and identifying clusters during the learning process. As a test case, we used the largest publicly available database of legal and illegal PAS consumption comprising 49, Colombian households. At the spatial level, these patterns indicate concentrations of drug consumption in specific regions of the country, which are closely related to specific geographic locations and the prevailing social and environmental contexts. These findings can provide valuable insights to facilitate decision-making and develop national policies targeting specific groups given their cultural, geographic, and social conditions. PAS can be grouped according to their chemical structure as synthetic cannabinoids, synthetic cathinones, phenethylamines, arylcyclohexylamines, tryptamines, indolalkylamines, new synthetic opioids, piperazines, ketamine, and designer benzodiazepines. They can also be grouped according to their origin as a natural origin or synthetic molecules \[ 2 — 4 \]. The increased numbers of drug use among young people are drawing the attention of national governments \[ 5 \]. Because the number of health-related incidents caused by using legal and illegal PAS worldwide has dramatically increased over the last two decades \[ 2 , 6 \], this phenomenon has become some of the largest burdens of disease \[ 7 , 8 \]. Drug use constitutes a high cost to society due to premature mortality, increased health expenditure, criminal justice drug and micro-trafficking , social welfare costs, and other social consequences \[ 9 , 10 \]. Colombia is ranked as one of the largest drug producers in the world \[ 11 \]. Unfortunately, the production and commercialization of drugs through drug- and micro-trafficking, constantly expands in locations with high levels of poverty and limited government presence \[ 12 \]. Furthermore, the country also has the highest prevalence of drug use among school students in recent years compared to other Latin American countries \[ 16 \]. The first step towards reducing this consumption is to identify drug consumption patterns in the general population \[ 17 \]. Several studies have identified patterns associated with drug use and consumption \[ 18 , 19 \]. According to the Center for Disease Control and Prevention, individuals who do not have their own homes and live in rented accommodations are more likely to use drugs \[ 20 \]. Other research studies suggest that neighborhood contextual characteristics may increase the risk of substance abuse \[ 21 — 26 \]. Additionally, population density may also influence substance use and overdose risk through a higher level of socialization in densely populated urban areas \[ 27 — 29 \]. Other authors have identified that anxiety, sleep disorders, suicide, depression, and other mental illnesses are risk factors for the consumption and abuse of PAS \[ 30 , 31 \]. Furthermore, early marijuana use has been shown to increase the risk of consuming other PAS \[ 32 \]. Furthermore, people involved in sports and artistic activities perceive drugs as enhancers element for improving their performance \[ 33 \]. In Colombia, drug consumption patterns and risk factors have also been identified. For instance, Kalyanam et al. Narvaez-Chicaiza \[ 35 \] assessed the social factors that lead to the adoption of harm reduction policies and how these factors influence treatments for substance abuse disorders. To our knowledge, we have not found any robust models integrating ML and spatial models to identify drug consumption patterns using publicly available Colombian databases. Although several techniques for analyzing drug consumption patterns are currently available i. On the other hand, logistic regression, confirmatory factor analysis, and correlational analysis are the most employed traditional statistical models to identify drug-associated patterns \[ 50 — 53 \]. Fraley and Raftery \[ 54 \] suggest separating clustering approaches into hierarchical and partitioning techniques. On the other hand, hierarchical techniques are divided into agglomerative and divisive methods. Although these techniques have been shown to perform well when relevant features are removed a priori , it is well-known that in clustering algorithms, irrelevant and redundant features in the data may degrade the quality of clusters and lead to high computational cost. Therefore, removing such features may alleviate these issues. Thus, we focus on identifying patterns of PAS consumption using an ensemble model integrating an autoencoder with both a clustering algorithm and a spatial model. As part of our approach, we used the most recent and representative works for data clustering, and different dimensionality reduction and feature selection methodologies proposed in the literature. Feature selection approaches in clustering can be split into filter, wrapper, embedded, and hybrid approaches \[ 37 \]. While wrappers depend on the clustering algorithms to evaluate the clustering quality of a selected feature subset, filters are independent of the clustering algorithm. Embedded approaches also work with a clustering algorithm and, unlike wrappers, incorporate knowledge about the clustering structure. Another type of method is hybrid approaches, which combine filter and wrapper approaches into a single strategy. However, studies on embedded and hybrid feature selection approaches in clustering are limited \[ 37 \]. Other feature learning-based approaches using Deep Neural Networks have been shown to work well for linear and nonlinear models \[ 56 \]. For instance, Xie et al. However, these are mainly used to work and process images. In general, deep clustering models use Auto-Encoders since they can learn input features without labels on the data; performance measures show that this approach is reliable for different data types \[ 58 \]. Thus, deep clustering methods have become a growing field of research for feature selection \[ 58 \]. In this regard, the use of convolutional networks in autoencoders and the application of feature selection for clustering are open questions that have not been fully addressed yet, especially when dealing with data from different statistical distributions \[ 37 \]. Here, we propose a Deep Neural Network-based Clustering-oriented Embedding Algorithm that allows us to i identify consumption patterns of PAS; and ii build an ensemble algorithm integrating an autoencoder with a clustering algorithm and a spatial model to deal with the feature space and cluster memberships. Our approach is based on the model proposed by Xie et al. Li et al. Our main hypothesis is that incorporating these two critical elements in our proposal will help to identify and better understand drug consumption patterns and support national policy development processes. Located in South America, the Republic of Colombia is a diverse country with a population of over 50 million people distributed over a territory of , square miles \[ 59 \], encompassing jungles, highlands, grasslands, deserts, coasts, and islands, distributed in six regions and 32 departments states \[ 60 \] See S1 Fig. It is worth noting that, unfortunately, Colombia has been a major producer of illegal drugs for a long time, which has had a significant impact on drug consumption and abuse. According to the United Nations Office on Drugs and Crime, Colombia is the first cocaine-producing country and the eighth country with the highest production of cannabis \[ 61 \]. In addition, the Colombian Drug Observatory indicates that the use of illicit substances in the territory has increased to Reports also indicate that consumption of licit substances such as alcohol and tobacco has recently increased dramatically \[ 62 \]. We used two databases to identify drug consumption patterns in Colombia. This survey includes observations of 49, households, where information on housing, location, general characteristics of individuals, consumption of legal and illegal PAS, and implemented treatments is registered. The second database comes from the Colombian Drug Observatory and contains information on the production of PAS per area during All these databases are fully available and completely anonymized. In this study, we used departments states as georeferenced areas using polygons i. Thus, an ethics statement approved by an ethics committee is not required since we are using public information without the identification or individual information of the people involved. However, unlike the Xie et al. In addition, a spectral clustering-based centroid estimation is proposed to achieve an improved initial centroid calculation. We chose the CAE-DEC framework based on its ability to reduce both the number of model parameters and the dimensionality, while creating clusters simultaneously. In our approach, an encoder structure is first applied to map the input vector into a lower feature space, called latent feature space LFS. Then, the LFS is independently passed through a decoder structure and a clustering layer to achieve an efficient clustering framework. The encoder-decoder combination DA attempts to extract a LFS preserving the relevant information from the original input data. On the other hand, the clustering layer seeks to execute an improved clustering assignment by minimizing the divergence between a target distribution and a centroid-based probability distribution. In the last stage of the framework, a spatial analysis was performed using the feature space generated from the autoencoder as input. Here, the spatial data exploration is initially performed using Global Spatial Autocorrelation to determine to which level the similarity between observations in a dataset relates to the similarity of the locations of such observations \[ 65 \]. We also measure the Local Spatial Autocorrelation, which focuses on the relationships between each observation and its surroundings, rather than providing a single-number summary of these relationships across the map \[ 69 \]. This is estimated based on the ability to determine whether spatial autocorrelation is present in a geographically referenced data set. Finally, we perform regionalization, which corresponds to a special kind of clustering where the objective is to group similar observations based on their statistical attributes and spatial location \[ 70 \]. In this sense, regionalization embeds the same logic as standard clustering techniques while applying a series of geographical constraints \[ 71 \]. The DA is a deep neural network architecture capable of learning unsupervised representations of an input data set. Typically, DA networks are used for dimensionality reduction or denoising tasks. The structure of a DA is based on two deep networks: a network to transform the original input data into a latent feature space, and a network trained to reconstruct the original input data using the extracted latent space as input. The first network, used to extract the latent space, is called the encoder, while the second is called the decoder. Compared to a DA, which is built with only fully connected layers, the CAE structure can reduce the number of parameters compared to a DA \[ 74 \]. The convolution operation can be denoted as:. For non-linear mapping, an activation function g. The clustering layer is inspired by Xie et al. Initially, a soft assignment is computed between the latent space, also known as embedded space, and the cluster centroids. Then, update steps are repeated to define the final cluster centroids and embedded space. The Kullback—Leibler KL divergence is used as loss function during the optimization procedure. The objective is to minimize de KL divergence between a soft clustering distribution Q and an auxiliar target distribution P. The KL loss is calculated as:. As in Xie et al. To compute the target distribution p km , the second power of q km is calculated, and a cluster normalization is applied as follows:. Then, by minimizing the divergence between P and Q , the embedding learning is achieved through highly confident assignments. As previously mentioned, the cluster centroids are initialized using a spectral clustering-based approach. The spectral clustering allows flexible distance metrics and provides better cluster estimations than K -means \[ 57 \]. However, most spectral clustering algorithms have high computational requirements. To overcome these computational requirements, random samples are taken to estimate the cluster centroids. As spectral clustering does not estimate any centroid during the learning process, once the clusters are defined, the mean of each cluster is used as the centroid estimator. Initially, the input data is normalized within the interval \[0, 1\]. This normalization allows the network to use the most advanced learning rate and avoid the vanishing gradient problems, as well as alleviate overfitting. Further, to achieve a better learning process, the last CONV layer in the decoder structure is activated by a sigmoid activation function. Firstly, a CAE model will be trained to minimize the reconstruction loss L r computed as. The total loss during this training step will be set as. The training process is shown in Table 1. The goal is to obtain a latent space that minimizes the total loss. Finally, the label of each embedded point is established as. A second database with PAS production figures, was used in the spatial analysis stage to correlate the PSA consumption and production. For evaluation and comparison purposes, we use the Calinski-Harabasz \[ 75 \], Davies-Bouldin \[ 76 \], and Silhouette \[ 77 \] index as intrinsic clustering metrics. Among all individuals, we identified three different clusters; Although the CAE model seeks to extract a LFS that preserves the essential characteristics of the input data, our proposed CAE-DEC model not only preserves these important characteristics but, at the same time, also forces the encoder structure to generate representative clusters while extracting the new feature space. It should be noted that the CAE model alone cannot determine the labels of each point or define clusters in the data. Thus, clusters in Fig 2 were obtained through spectral clustering and were the bases for initializing the centroids in the CAE-DEC model. Out of the individuals in the sample, only Our results indicate that individuals in clusters 0 and 2 are more likely to consume some PAS Fig 3a , while most individuals in cluster 1 do not Table 2. In particular, Table 3 shows the adjusted residuals for our model. According to our results, the Central-Eastern region significantly contributes to the Region variable. In this region, the observed value is higher than the expected value in cluster 2, while the observed value is lower than the expected value for cluster 0. Indeed, this region shows fewer observed individuals than the expected number of individuals in cluster 2 and a higher number observed than expected individuals in clusters 0 and 1 Table 3. On the other hand, Gender has a higher-than-expected value of males in clusters 0 and 2, while it is lower in cluster 1. For females, the opposite occurs in cluster 1, and lower values are observed in clusters 0 and 2. Similarly, Housing Type has a higher-than-expected value of individuals living at houses in cluster 1 and a lower-than-expected in cluster 2. Conversely, cluster 2 has more individuals living in apartments, and cluster 1 has the lowest Table 3. Regarding SES, a higher-than-expected number of individuals in strata 3, 4, 5, and 6 in cluster 0 were found Table 3. We also observed a lower-than-expected number of individuals in strata 3, 4, 5, and 6 in cluster 2 and a higher-than-expected number in strata 1, 4 and 6 in cluster 1 Table 3. Moreover, the age variable shows a higher-than-expected observed value for the 0,20\] range in cluster 0. For ages between 20,40\] years, cluster 2 has a higher-than-expected number of individuals. Conversely, there is a lower number of individuals in cluster 1. Finally, the household economy variable results show that cluster 2 has a higher-than-expected value of individuals contributing to the household finances, and cluster 1 has a lower-than-expected value of individuals not contributing to it. Different alternative classification algorithms were used to determine the number of choropleth class limits i. According to our results, the Fisher-Jenks classifier performed better and hence was selected. Following the same exploratory spatial analysis, we constructed a choropleth with the percentage of PAS use for each of the 32 Colombian departments Fig 5a. However, some of these departments are major drug producers i. Here, HH, LH, LL and ns represent high-high, low-high, low-low, and not statistically significant quadrants, respectively. Thus, the null hypothesis that the map is random i. Specifically, the high-high HH and low-low LL quadrants indicate a positive association between high and low drug use. On the other hand, the low-high LH and high-low HL quadrants indicate negative associations with drug use Fig 5b. Thus, a little over We also identified that, among legal drugs, alcohol and tobacco are the most frequently consumed in the national territory Fig 6a. It should be mentioned that the use of these drugs is also present across the country but with a lower incidence Fig 7. For interpretation purposes, number represents values scaled on a range of 0 to 1. Concerning illegal drugs, non-prescription tranquilizers and stimulants are most prevalent in Casanare Fig 8. As for cocaine, its consumption is the highest in Risaralda and moderately high in Antioquia Fig 8. Number represents values scaled on a range of 0 to 1 for psychoactive substance use. Conventions as in Fig 7. On the other hand, basuco i. Finally, 2CB has the highest consumption rate in Risaralda, followed by Caldas. Although the consumption pattern of some departments is not mentioned, there is low and moderate consumption for certain drugs in some of them Fig 8. We applied a regionalization method as a grouping technique for imposing a spatial restriction, i. Our approach uses a spatially constrained hierarchical clustering algorithm, which identified three clusters representing the consumption of PAS in the country Fig 9. The number of clusters was estimated based on the average silhouette indexes, the total intra-cluster variance, and dendrograms S2 Fig. In addition, the feature coherence i. In this study, we propose and test a Deep Neural Network-based Clustering-oriented Embedding algorithm i. This model allows the automatic extraction of features from the input data such as sex, age, socioeconomic status, and housing type to determine whether an individual has consumed PAS. It then creates clusters in the new data space generated during the learning process, following the methods outlined in \[ 56 , 57 \]. After the training process, a latent feature space LFS is generated, and the results are subsequently analysed. We have identified clearly marked clusters where the prevalence of individuals who use or do not use PAS is notable. Additionally, we found that region, sex, housing type, socioeconomic strata, age, and whether individuals contribute to household finances have a statistically significant impact on the clustering structure. These findings are consistent with previous studies aimed at identifying PAS consumption patterns \[ 19 , 79 , 80 \]. Based on our findings, individuals more likely to consume PAS are grouped in cluster 2, while cluster 1 consisted of individuals who did not consume PAS Table 2. Not surprisingly, a significant proportion of females characterizes cluster 1. In addition, most individuals belong to socioeconomic strata 1, are 40 years old or older, and do not contribute economically to support their household. In contrast, cluster 2 is characterized by a higher proportion of males aged between 20 and 40 in socioeconomical strata 1 and 2, who do not contribute to the household finances Table 2. Finally, cluster 0 is characterized by a small proportion of males, a higher proportion of individuals in strata 3, 4, 5, and 6, and individuals are more likely to contribute to the household economy Table 2. At the level of spatial statistics, we identified that legal drugs such as alcohol have a high prevalence in all regions of Colombia, with a slight tendency to more consumption in coastal areas Fig 7. In our country, the coastal areas are often popular tourist destinations, and many tourists come to these areas looking for a relaxing experience, which can increase alcohol consumption. Coastal areas typically have warmer temperatures and more sunshine, increasing thirst and making people more likely to consume beverage. Additionally, bars, clubs, and restaurants serve alcoholic beverage due to the high demand from tourists and locals \[ 81 , 82 \]. Another characteristic of this area is the fishing and maritime culture. This culture is often associated with hard work and long working hours, and alcohol may be seen as a way to relax and unwind after a tough day at the sea \[ 83 \]. The level of development, as measured by gross domestic product GDP , is the third region with significant economic development in the country \[ 84 \] S3 Table. Interestingly, the consumption of illegal drugs is lower in the Northern region than in other regions of the country. However, there is a more representative consumption of non-prescription tranquilizers, opioids, ketamine, GHB, and heroin. Tobacco consumption is present in all regions, with a higher proportion in the Central region Eje Cafetero—Antioquia , where climate conditions resemble temperate weather. Also, this region has a diverse consumption pattern, where drugs such as marijuana, popper, cocaine, ecstasy, inhalants, methadone, heroin, LSD, GHB, 2CB, and mushrooms prevail. Energy drinks are more frequently used in the Central-Eastern region, characterized by a continental climate surrounded by flat territory. Our results are in line with the scientific literature suggesting that the location of regions within countries is directly associated with the consumption of PAS \[ 26 , 85 — 87 \]. The consumption of heroin, basuco, non-prescription tranquilizers, stimulants, methamphetamines, opioids, and ketamine characterizes this region. One of the main reasons for this result is that, unfortunately, this region has favourable environmental characteristics i. In the Western region Pacific , also known as the Pacific region, consumption mostly mainly includes of Methylene Chloride, GHB, heroin, opioids, and methamphetamines. This region Pacific is mainly characterized known for its geographical isolation, poverty, and ongoing conflict, which have contributed to the growth of drug production and trafficking in the area. Poverty is one of the main factors driving drug production in the Pacific region, which has led many people to turn to drug cultivation and trafficking for survival. In the Western region, also known as the Pacific region, consumption mainly includes methylene chloride, GHB, heroin, opioids, and methamphetamines. This region is mainly characterized for its geographical isolation, poverty, and ongoing conflict, which have contributed to the growth of drug production and trafficking in the area. On the other hand, this region ranks second among the regions with the lowest levels of development S3 Table. In summary, the proposed CAE-DEC model simultaneously integrates a feature extraction process within the clustering design, prioritizing features that improve the separation between groups, thus avoiding the manual extraction of features, which is a frequent process in traditional models. Additionally, a geospatial component is sequentially included to expand the resulting insights by considering geographic constraints. Currently, these types of architectures are scarce in understanding mental health problems. As part of future work, the architecture of the proposed model could be improved to integrate the automatic extraction of features while optimizing a geospatial loss. Following our experience with the proposed CAE-DEC in PAS consumption, the application of this model to other mental health problems, such as suicide, depression, and domestic violence, among other pathologies, could be explored. This can include education, treatment, and harm reduction programs. Also, this information can be used to develop public health campaigns to raise awareness about the risks of drug use and reduce their negative impact. Furthermore, this information can be used to crack down on drug trafficking and distribution networks. On the other hand, this information can be used to alert healthcare providers and regulatory bodies to take appropriate action to prevent their use and discover new drugs. Some of this work is to be presented to the Ph. The data used in the manuscript were obtained from a third party, the Archivo Nacional de Datos ANDA , and are fully available and anonymized. The authors confirm that others would be able to access these data in the same manner as themselves; and the authors did not have any special access privileges that others would not have. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process. Please submit your revised manuscript by Apr 02 PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone plos. A rebuttal letter that responds to each point raised by the academic editor and reviewer s. You should upload this letter as a separate file labeled 'Response to Reviewers'. A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'. An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'. If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter. If applicable, we recommend that you deposit your laboratory protocols in protocols. Please confirm that all data sources you used were publicly available and anonymized. Furthermore, please clarify how the data were accessed for the purpose of this study. List the grants or organizations that supported your study, including funding received from your institution. Please include your amended statements within your cover letter; we will change the online submission form on your behalf. The sponsor of the study has no role in study design; in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the article for publication. We note that you have provided funding information that is not currently declared in your Funding Statement. However, funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form. Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. Any potentially identifying patient information must be fully anonymized. Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access. We will update your Data Availability statement to reflect the information you provide in your cover letter. We note that Figures 1, 5 9 and S2 in your submission contain map images which may be copyrighted. For these reasons, we cannot publish previously copyrighted maps or satellite images created using proprietary data, such as Google software Google Maps, Street View, and Earth. We require you to either 1 present written permission from the copyright holder to publish these figures specifically under the CC BY 4. You may seek permission from the original copyright holder of Figures 1, 5, 9 and S2 to publish the content specifically under the CC BY 4. Please be aware that this license allows unrestricted use and distribution, even commercially, by third parties. Please upload the completed Content Permission Form or other proof of granted permissions as an ''Other'' file with your submission. If you are unable to obtain permission from the original copyright holder to publish these figures under the CC BY 4. Please check copyright information on all replacement figures and update the figure caption with source information. If applicable, please specify in the figure caption text when a figure is similar but not identical to the original image and is therefore for illustrative purposes only. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented. The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception please refer to the Data Availability Statement in the manuscript PDF file. The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e. PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here. Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. Please upload your review as an attachment if it exceeds 20, characters. Reviewer 1: The manuscript was well constructed. However, the methodology seems very technical, with no explanation of what data were used, how and where the analyzes were carried out. I suggest making the methodology clearer, for example in CAE not only citing authors who used the technique, but making it clear why it was chosen and not another deep learning technique. In the results, improve the identification of tables and figures. Include subtitles to make files more understandable. In the discussion, explore further the impact of the results and assess the environmental issue of these sites. There was a stratification of regions and consumption profile, but there is no information about the location. The area is urban or rural, what level of development. PLOS authors have the option to publish the peer review history of their article what does this mean? If published, this will include your full peer review and any attached files. Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy. Please log into your account, locate the manuscript record, and check for the action link 'View Attachments'. If this link does not appear, there are no attachment files. To use PACE, you must first register as a user. Registration is free. Please note that Supporting Information files do not need this step. Please find below our response in blue. We very much appreciate all your efforts as Editor-in-Chief and the detailed and extraordinary revision this manuscript had. We enjoyed the thorough review of our manuscript, and it was a pleasure to respond to the reviewers. Please, Dr. Silva, allows us to mention that this exercise of the fair and exigent peer-review process is disappearing, and only good journals like PLOS ONE keep it. We very much appreciate your comments. In particular,. Thank you very much for your comments. For practicality, we downloaded the data and stored it on the cloud. Thank you for your comment. As per Colombian regulations, the data were fully anonymized by the Colombian government before we could use them. However, Universidad del Norte did not provide any additional funds for this study. The following statement was included in the revised version of the manuscript to address this:. We appreciate your correction. As per your request, we have removed any funding-related text from the manuscript and clarified that Universidad del Norte provided only tuition expenses for our PhD student. According to the author, this file can be freely accessible and used, which indicates that neither a license nor permission is needed to use it i. Kindly see our response to Comment 6. As mentioned above, a license to use the maps is not required. We very much appreciate your supporting comments. We changed the Supporting Information citation on the manuscript i. The manuscript was well constructed. I suggest making the methodology clearer, for example in CAE not only citing authors who used the technique but making it clear why it was chosen and not another deep learning technique. All notebooks and code written for processing and analyzing the data are available from first author under reasonable request. For evaluation and comparison purposes, we use the Calinski-Harabasz \[76\], Davies-Bouldin \[77\], and Silhouette \[78\] index as intrinsic clustering metrics. On the other hand, we rewrote a paragraph in the methodology to make clear why the model was chosen. Now, the text reads:. In the Discussion, explore further the impact of the results and assess the environmental issue of these sites. Thank you for your suggestion. We improved the discussion section accordingly in the revised version of the manuscript. The relevant text now reads:. It then creates clusters in the new data space generated during the learning process, following the methods outlined in \[57, 59\]. After the training process, a latent feature space LFS is generated and the results are subsequently analysed. These findings are consistent with previous studies aimed at identifying PAS consumption patterns \[19, 80, 81\]. Based on our findings, individuals more likely to consume PAS are grouped in cluster 2, while cluster 1 consisted of individuals who did not consume PAS Table 1. In contrast, cluster 2 is characterized by a higher proportion of males aged between 20 and 40 in socioeconomical strata 1 and 2, who do not contribute to the household finances Table 1. Finally, cluster 0 is characterized by a small proportion of males, a higher proportion of individuals in strata 3, 4, 5, and 6, and individuals are more likely to contribute to the household economy Table 1. Coastal areas typically have warmer temperatures and more sunshine, increasing thirst and making people more likely to consume alcohol. Additionally, bars, clubs, and restaurants serve alcoholic beverage due to the high demand from tourists and locals \[82, 83\]. This culture is often associated with hard work and long working hours, and alcohol may be seen as a way to relax and unwind after a tough day at the sea \[84\]. The level of development, as measured by gross domestic product GDP , is the third region with significant economic development in the country \[85\] S3 Table, Supplementary Material. Tobacco consumption is present in all regions, with a higher proportion in the Central region, where climate conditions resemble temperate weather. Energy drinks are more frequently used in the Eastern region, characterized by a continental climate surrounded by flat territory. Our results are in line with the scientific literature suggesting that the location of regions within countries is directly associated with the consumption of PAS \[26, \[86—88\]. On the other hand, this region ranks second among the regions with the lowest levels of development S3 Table, Supplementary Material. In this sense, feature selection approaches can be useful for classification, clustering, or regression. In the Introduction, please give problem, challenges clearly. Introduction is written separately. Thank you for your input. Following your advice, several changes have been made in the Introduction of the revised version of the manuscript to address this. Both Convolutional Autoencoder CAE and Stacked Autoencoder SAE are types of autoencoders, a type of neural network architecture that is used for unsupervised learning and data compression. A CAE is typically used for processing image data. It uses convolutional layers to extract spatial features from the input image and then uses deconvolutional layers to reconstruct the image. CAEs are well-suited for image data because they can capture the spatial relationships between pixels in an image and can learn to recognize visual patterns and shapes. On the other hand, SAE is typically used for processing structured or unstructured data. It consists of multiple layers of neural networks that encode the input data into a lower-dimensional representation and then decode it back to the original dimensions. SAEs are useful for feature learning and data compression in many different types of data, including text, audio, and structured data. In summary, while CAE is focused on processing image data using convolutional layers, SAE can be applied to various types of data and uses multiple layers of neural networks for encoding and decoding. Thank you for the comments. On the other hand, our model has the smallest Davies-Bouldin score, which indicates that identified clusters groups have a better partition. Regarding the Silhouette index, our model gives the highest value, which implies that clusters are highly dense. Silhouette index: The best value is 1 and the worst value is Values near 0 indicate overlapping clusters. Negative values generally indicate that a sample has been assigned to the wrong cluster, as a different cluster is more similar. The problem we are addressing is this study is a clustering problem, as we aim to identify consumption patterns of psychoactive substances PAS in the Colombian territory. In particular, we used clustering techniques i. More information about performance metrics can be found in S2 Table of the Supplementary Material. According to our results, the proposed CAE-DEC model shows a well-separated and highly dense cluster, meaning we can define better groups and identify PAS consumer patterns more precisely. Please submit your revised manuscript by Aug 25 PM. Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. I suggest including a location map in the methodology section. The map will offer a clear spatial reference, enabling readers to visualize precise locations of study points or mentioned areas. The inclusion of a location map can improve overall clarity and comprehension, making the article more accessible to a broad audience, including non-specialist readers. We have added a locations map in the methodology section. Additionally, bars, clubs, and restaurants serve alcoholic beverage due to the high demand from tourists and locals. An invoice for payment will follow shortly after the formal acceptance. If you have any billing related questions, please contact our Author Billing department directly at authorbilling plos. If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress plos. Your manuscript is now with our production department. If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. For more information please contact onepress plos. This section collects any data citations, data availability statements, or supplementary materials included in this article. As a library, NLM provides access to scientific literature. PLoS One. Leading consumption patterns of psychoactive substances in Colombia: A deep neural network-based clustering-oriented embedding approach Kevin Palomino Kevin Palomino 1 Department of Industrial Engineering, Universidad del Norte, Barranquilla, Colombia. Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Visualization, Writing — original draft. Find articles by Kevin Palomino. Find articles by Carmen R Berdugo. Received Oct 27; Accepted Jul 22; Collection date Open in a new tab. Training process: 1. Generate an initial latent space Z through the pre-trained CAE 2. Run spectral clustering with Z to generate the initial cluster centers C 3. Weight, Bias, and Centers. Output: Latent space, labels. Click here for additional data file. S3 Table. Characteristics of the level of development, urbanity, rurality, and drug production in the regions of Colombia. S1 Fig. Location map. S2 Fig. The optimal number of clusters using dendrogram. S3 Fig. PMC Copyright notice. Attachment Submitted filename: Response to Reviewers. Associated Data. Similar articles. Add to Collections. Create a new collection. Add to an existing collection. Choose a collection Unable to load your collection due to an error Please try again. Add Cancel. Generate an initial latent space Z through the pre-trained CAE. Run spectral clustering with Z to generate the initial cluster centers C. Calculate soft assignment distribution Q and target distribution P based on Z and C. Obtain the label for each data point from the las optimized Q.

Buy Ecstasy online in Santa Marta