Unsupervised Classification Model Research Articles

Automated patch correctness predicting to fix software defect

Automated Program Repair (APR) is a technique that can automatically fix software defects without manual debugging, playing a crucial role in software development and maintenance. However, the patches generated by APR still suffer from the problem of overfitting, which poses a significant threat to practical applications. Previous studies have proposed various approaches to predict the correctness of the patch to address this issue, primarily including static-based methods, dynamic-based methods, and learning-based methods. However, these methods all have their own limitations, making it difficult to accurately extract code semantic features and achieve comprehensive prediction. To address the aforementioned challenges, we propose a learning-based unsupervised classification model, Automated Patch cOrrectness aSsessmenT based on muLtiple pErspectives (APOSTLE), for predicting the correctness of patches. Specifically, APOSTLE consists of three components: code vectorization component, where advanced pre-trained models are used to achieve efficient extraction of semantic features from the code; similarity and code change degree calculation component, where APOSTLE calculates the similarity and the degree of change to the code; comprehensive evaluation component, where APOSTLE conducts comprehensive evaluation, addressing the issue of prediction comprehensiveness. Experiments on a collection of 1278 patches (written by developers or generated by 32 APR tools) demonstrate that APOSTLE achieves an AUC value of 0.801, an MAP value of 0.855, and an MRR value of 0.944, outperforming the state-of-the-art approach BATS by 8.3%, 6.0%, and 9.0%, respectively. APOSTLE successfully achieves accurate extraction of code semantic features while achieving comprehensive patch correctness prediction.

A novel breast cancer image classification model based on multiscale texture feature analysis and dynamic learning

Assistive medical image classifiers can greatly reduce the workload of medical personnel. However, traditional machine learning methods require large amounts of well-labeled data and long learning times to solve medical image classification problems, which can lead to high training costs and poor applicability. To address this problem, a novel unsupervised breast cancer image classification model based on multiscale texture analysis and a dynamic learning strategy for mammograms is proposed in this paper. First, a gray-level cooccurrence matrix and Tamura coarseness are used to transfer images to multiscale texture feature vectors. Then, an unsupervised dynamic learning mechanism is used to classify these vectors. In the simulation experiments with a resolution of 40 pixels, the accuracy, precision, F1-score and AUC of the proposed method reach 91.500%, 92.780%, 91.370%, and 91.500%, respectively. The experimental results show that the proposed method can provide an effective reference for breast cancer diagnosis.

Machine learning-based quantitative analysis of metal ductile fracture surface

Quantitative fracture analysis (QFA) can infer mechanical properties and fracture mechanism of materials based on quantitative statistics of various topographical features on the fracture surface. However, it is still a challenge to screen the key fracture surface descriptors for a given property through experimental insight and theory. Obtaining an accurate quantitative relationship is difficult as different properties involve a variety of fracture surface descriptors. Herein, a QFA based on machine learning is proposed. The fracture surface images are first divided into three classes: intergranular, quasi-cleavage and dimple, by an unsupervised classification model based on a convolutional neural network with an accuracy of 0.867. The fracture surface descriptors of the identified dimple images are extracted by image segmentation and used to train separate regression models to predict the values of fracture toughness (KIC), tensile strength (Rm), yield strength (Re), elongation after fracture (A) and fracture shrinkage (Z). These models demonstrate good performance with an R2 in the test set of 0.934, 0.944, 0.840, 0.890, 0.900, respectively. In addition, with KIC and Rm as dependent variables, the explicit relationships between them and key morphological features are established based on symbolic regression. The R2 between the predicted and experimental values of the two formulae reach 0.925 and 0.926, respectively. Such a machine learning-based QFA can assist in the rapid prediction and evaluation of mechanical properties of materials in engineering applications.

A traffic anomaly detection approach based on unsupervised learning for industrial cyber–physical system

In Industrial Cyber–Physical Systems (ICPSs), the attacker can intrude into the cyber system through many penetration tools and attack the physical system. Payload-based traffic anomaly detection is a popular technique against these attacks. Due to the imbalanced distribution of normal and attack samples in ICPS, existing payload-based detection methods are mostly implemented based on unsupervised learning, typically comprising a word segmentation model and an unsupervised classifier. However, existing methods may disrupt semantic correlations and face challenges in extracting complex payload dependence relationships. To address these issues, this paper proposes a traffic anomaly detection approach, which consists of a data preprocessing model, an unsupervised word segmentation model, and an unsupervised classification model based on autoencoder. The unsupervised word segmentation model utilizes Long Short-Term Memory (LSTM) to calculate the probability of each word segmentation combination, effectively addressing the issue of inaccurate segmentation results in existing payload segmentation models. The unsupervised classification model, which combines 1D-Convolutional Neural Network (1D-CNN) and Bidirectional Encoder Representation from Transformers (BERT), addresses the challenge of extracting complex payload dependence relationships in existing classification models. The proposed detection approach is evaluated using a Cyber–Physical Attack Dataset (CPAD). Compared with the state-of-the-art detection approaches, the proposed approach has shown a significant improvement in Precision, with an increase of 18.83%. Additionally, the Recall has also been substantially enhanced, with a gain of 22.3%. Overall, the F1 has demonstrated a comprehensive improvement of 20.60%.

Game theory and MCDM-based unsupervised sentiment analysis of restaurant reviews.

Sentiment Analysis is a method to identify, extract, and quantify people's feelings, opinions, or attitudes. The wealth of online data motivates organizations to keep tabs on customers' opinions and feelings by turning to sentiment analysis tasks. Along with the sentiment analysis, the emotion analysis of written reviews is also essential to improve customer satisfaction with restaurant service. Due to the availability of massive online data, various computerized methods are proposed in the literature to decipher text sentiments. The majority of current methods rely on machine learning, which necessitates the pre-training of large datasets and incurs substantial space and time complexity. To address this issue, we propose a novel unsupervised sentiment classification model. This study presents an unsupervised mathematical optimization framework to perform sentiment and emotion analysis of reviews. The proposed model performs two tasks. First, it identifies a review's positive and negative sentiment polarities, and second, it determines customer satisfaction as either satisfactory or unsatisfactory based on a review. The framework consists of two stages. In the first stage, each review's context, rating, and emotion scores are combined to generate performance scores. In the second stage, we apply a non-cooperative game on performance scores and achieve Nash Equilibrium. The output from this step is the deduced sentiment of the review and the customer's satisfaction feedback. The experiments were performed on two restaurant review datasets and achieved state-of-the-art results. We validated and established the significance of the results through statistical analysis. The proposed model is domain and language-independent. The proposed model ensures rational and consistent results.

Mapping Phonation Types by Clustering of Multiple Metrics

For voice analysis, much work has been undertaken with a multitude of acoustic and electroglottographic metrics. However, few of these have proven to be robustly correlated with physical and physiological phenomena. In particular, all metrics are affected by the fundamental frequency and sound level, making voice assessment sensitive to the recording protocol. It was investigated whether combinations of metrics, acquired over voice maps rather than with individual sustained vowels, can offer a more functional and comprehensive interpretation. For this descriptive, retrospective study, 13 men, 13 women, and 22 children were instructed to phonate on /a/ over their full voice range. Six acoustic and EGG signal features were obtained for every phonatory cycle. An unsupervised voice classification model created feature clusters, which were then displayed on voice maps. It was found that the feature clusters may be readily interpreted in terms of phonation types. For example, the typical intense voice has a high peak EGG derivative, a relatively high contact quotient, low EGG cycle-rate entropy, and a high cepstral peak prominence in the voice signal, all represented by one cluster centroid that is mapped to a given color. In a transition region between the non-contacting and contacting of the vocal folds, the combination of metrics shows a low contact quotient and relatively high entropy, which can be mapped to a different color. Based on this data set, male phonation types could be clustered into up to six categories and female and child types into four. Combining acoustic and EGG metrics resolved more categories than either kind on their own. The inter- and intra-participant distributional features are discussed.

Unsupervised classification model of residential power users based on typical load patterns

It is critical for the smart grid to mining the power users’ consumption behaviors. The accumulation of power data provides the possibility of conducting experiments for research focus on behavior analysis. This paper proposes a residential power users’ classification method based on typical load patterns called UCM-LP, to classify users’ power consumption behavior. The method firstly performs fuzzy grouping on the typical power consumption characteristics of multiple users through two-stage clustering, then divides cluster label to obtain the final power consumption categories of the user. Finally, this paper verified the algorithm using the real household power consumption data, obtaining the specific power consumption category patterns and discussed the differences among power consumption categories.

Unsupervised domain adaptive myocardial infarction MRI classification diagnostics model based on target domain confused sample resampling

ObjectiveInefficient circulatory system due to blockage of blood vessels leads to myocardial infarction and acute blockage. Myocardial infarction is frequently classified and diagnosed in medical treatment using MRI, yet this method is ineffective and prone to error. As a result, there are several implementation scenarios and clinical significance for employing deep learning to develop computer-aided algorithms to aid cardiologists in the routine examination of cardiac MRI. MethodsThis research uses two distinct domain classifiers to address this issue and achieve domain adaptation between the particular field and the specific part is a problem Current research on environment adaptive systems cannot effectively obtain and apply classification information for unsupervised scenes of target domain images. Insufficient information interchange between specific domains and specific domains is a problem. In this study, two different domain classifiers are used to solve this problem and achieve domain adaption. To effectively mine the source domain images for classification understanding, an unsupervised MRI classification technique for myocardial infarction called CardiacCN is proposed, which relies on adversarial instructions related to the interpolation of confusion specimens in the target domain for the conflict of confusion specimens for the target domain classification task. ResultsThe experimental results demonstrate that the CardiacCN model in this study performs better on the six domain adaption tasks of the Sunnybrook Cardiac Dataset (SCD) dataset and increases the mean target area myocardial infarction MRI classification accuracy by approximately 1.2 percent. The classification performance of the CardiacCN model on the target domain does not vary noticeably when the temperature-controlled duration hyper-parameter rl falls in the region of 5–30. According to the experimental findings, the CardiacCN model is more resistant to the excitable rl. The CardiacCN model suggested in this research may successfully increase the accuracy of the source domain predictor for the target domain myocardial infarction clinical scanning classification in unsupervised learning, as shown by the visualization analysis infrastructure provision nurture. It is evident from the visualization assessment of embedded features that the CardiacCN model may significantly increase the source domain classifier's accuracy for the target domain's classification of myocardial infarction in clinical scans under unsupervised conditions. ConclusionTo address misleading specimens with the inconsistent classification of target-domain myocardial infarction medical scans, this paper introduces the CardiacCN unsupervised domain adaptive MRI classification model, which relies on adversarial learning associated with resampling target-domain confusion samples. With this technique, implicit image classification information from the target domain is fully utilized, knowledge transfer from the target domain to the specific domain is encouraged, and the classification effect of the myocardial ischemia medical scan is improved in the target domain of the unsupervised scene.

Thrombosis origin identification of cardioembolism and large artery atherosclerosis by distinct metabolites

BackgroundThe diagnosis of cerebral thrombosis origin is challenging and remains unclear. This study aims to identify thrombosis due to cardioembolism (CE) and large artery atherosclerosis (LAA) from a new perspective...

DF-SSmVEP: Dual Frequency Aggregated Steady-State Motion Visual Evoked Potential Design with Bifold Canonical Correlation Analysis.

Recent advancements in Electroencephalographic (EEG) sensor technologies and signal processing algorithms have paved the way for further evolution of Brain Computer Interfaces (BCI) in several practical applications, ranging from rehabilitation systems to smart consumer technologies. When it comes to Signal Processing (SP) for BCI, there has been a surge of interest on Steady-State motion Visual Evoked Potentials (SSmVEP), where motion stimulation is used to address key issues associated with conventional light flashing/flickering. Such benefits, however, come with the price of being less accurate and having a lower Information Transfer Rate (ITR). From this perspective, this paper focuses on the design of a novel SSmVEP paradigm without using resources such as trial time, phase, and/or number of targets to enhance the ITR. The proposed design is based on the intuitively pleasing idea of integrating more than one motion within a single SSmVEP target stimuli, simultaneously. To elicit SSmVEP, we designed a novel and innovative dual frequency aggregated modulation paradigm, called the Dual Frequency Aggregated Steady-State motion Visual Evoked Potential (DF-SSmVEP), by concurrently integrating “Radial Zoom” and “Rotation” motions in a single target without increasing the trial length. Compared to conventional SSmVEPs, the proposed DF-SSmVEP framework consists of two motion modes integrated and shown simultaneously each modulated by a specific target frequency. The paper also develops a specific unsupervised classification model, referred to as the Bifold Canonical Correlation Analysis (BCCA), based on two motion frequencies per target. The corresponding covariance coefficients are used as extra features improving the classification accuracy. The proposed DF-SSmVEP is evaluated based on a real EEG dataset and the results corroborate its superiority. The proposed DF-SSmVEP outperforms its counterparts and achieved an average ITR of 30.7 ± 1.97 and an average accuracy of 92.5 ± 2.04, while the Radial Zoom and Rotation result in average ITRs of 18.35 ± 1 and 20.52 ± 2.5, and average accuracies of 68.12 ± 3.5 and 77.5 ± 3.5, respectively.

Noninvasive Diagnostic for COVID-19 from Saliva Biofluid via FTIR Spectroscopy and Multivariate Analysis.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has caused the worst global health crisis in living memory. The reverse transcription polymerase chain reaction (RT-qPCR) is considered the gold standard diagnostic method, but it exhibits limitations in the face of enormous demands. We evaluated a mid-infrared (MIR) data set of 237 saliva samples obtained from symptomatic patients (138 COVID-19 infections diagnosed via RT-qPCR). MIR spectra were evaluated via unsupervised random forest (URF) and classification models. Linear discriminant analysis (LDA) was applied following the genetic algorithm (GA-LDA), successive projection algorithm (SPA-LDA), partial least squares (PLS-DA), and a combination of dimension reduction and variable selection methods by particle swarm optimization (PSO-PLS-DA). Additionally, a consensus class was used. URF models can identify structures even in highly complex data. Individual models performed well, but the consensus class improved the validation performance to 85% accuracy, 93% sensitivity, 83% specificity, and a Matthew’s correlation coefficient value of 0.69, with information at different spectral regions. Therefore, through this unsupervised and supervised framework methodology, it is possible to better highlight the spectral regions associated with positive samples, including lipid (∼1700 cm–1), protein (∼1400 cm–1), and nucleic acid (∼1200–950 cm–1) regions. This methodology presents an important tool for a fast, noninvasive diagnostic technique, reducing costs and allowing for risk reduction strategies.

Exploratory Data Analysis of Community Behavior Towards the Generation of Solid Waste Using K-Means and Social Indicators

In Peru, solid waste accumulation has been constant for decades and impacts 72% of local governments, affecting 42% of the population. These numbers show new tools are required to better understand this phenomenon and develop appropriate mitigation methods. In this light, this research proposes an exploratory analysis of the study population against the accumulation of solid waste. For this, the study proposes the segmentation of a specific population through a set of social indicators grouped into three categories of analysis (i.e., sociocultural, sociodemographic, and socioeconomic) and, in turn, assess the geographic proximity between each group of people segmented according to the parameters used for this study, and the informal points of accumulation of MSW. To segment the study population, an unsupervised classification model (i.e., K-means) was used. For methodological purposes, the Puente Piedra district was chosen as a case study. The results show that the predominant population is framed between the ages of 36 to 45, with an intermediate educational level (i.e., secondary school) and an approximate monthly income of $ 300. In addition, the predominant family structure includes up to four members living in the same household. Finally, it is observed that the behavior of people who live close as neighbors is similar and is also related to the geographic location of the dumps.

Proposal of Unsupervised Gas Classification by Multimode Microresonator

A high-accuracy unsupervised classification model is developed in a multimode self-interference microring resonator (SIMRR). For the SIMRR, there are many whispering gallery modes (WGMs) present. Each of these resonance modes supported by the SIMRR has a different response to different target parameters, so that the SIMRR-based sensor has the super capability to distinguish between multiple components. In the classification model, principal component analysis (PCA) is firstly used to reduce the dimensionality of this multimode sensing information from the SIMRR-based sensor. When the original higher-dimensional data points are projected onto the lower-dimensional data with only the first few principal components, they can be easily categorized into several different types by using density-based spatial clustering of application with noise (DBSCAN) algorithm. As an example, the unsupervised classification method is numerically validated based on a designed three-gas SIMRR-based sensor. The numerical results prove that the classification model can achieve an ultra high classification accuracy for the designed three-gas sensor with more than 60 dB in signal-to-noise ratio.

Unsupervised classification of specialty coffees in Homogeneous sensory attributes through machine learning

Brazil is the largest exporter of coffee beans, 29% world exports, 15% this volume in specialty coffees. Thereby researches are done, so that identify different segments in the market, in order to direct the end consumer to a better quality product. New technologies are explored to meet an increasing demand for high quality coffees. Therefore, in this article has an objective to propose the use of machine learning techniques combined with projection pursuit in the construction of unsupervised classification models, in a sensory acceptance experiment, applied to four groups of trained and untrained consumers, in four classes of specialty coffees in which they were evaluated sensory characteristics: aroma, body coffee, sweetness and general note. For evaluating classifier performance, in the data with reduced dimension, all instances were used, and considering four groupings, the models were adjusted. The results obtained from the groupings formed were compared with pre-established classes to confirm the model. Success and error rates were obtained, considering the rate of false positives and false negatives, sensitivity and classification methods accuracy. It was concluded that, machine learning use in data with reduced dimensions is feasible, as it allows unsupervised classification of specialty coffees, produced at different altitudes and processes, considering the heterogeneity among consumers involved in sensory analysis, and the high homogeneity of sensory attributes among the analyzed classes, obtaining good hit rates in some classifiers. Key words: Classification models; Data dimension reduction; Groupings identification; Projection pursuit.

Catchment-scale, high-resolution, hydraulic models and habitat maps – a salmonid's perspective

The advent of remotely-sensed high-resolution imagery has led to the development of methods to map river bathymetry. In this study, we utilized high-resolution imagery to map river depth and quantify hydraulic habitats at the catchment scale (>1000 km2) during low flows. Using 0.3-m airborne multi-spectral imagery (resampled to 0.5 m), we mapped contiguous river depth (124 km) within a well-established Atlantic Salmon (Salmo salar) and Brook Trout (Salvelinus fontinalis) river – The Little Southwest Miramichi, New Brunswick. We built image-derived depth maps with and without field data calibration. The model without field calibration data (flow resistance equation‐based imaging of river depths) accurately described river depths (R 2 = 72.7; RMSE = 0.167 m; n = 762); however, it overestimated shallow depths. The field-calibrated model removed shallow depth errors (R 2 = 76.4; RMSE = 0.155 m; n = 762). We mapped velocity using a relationship between river geometry and discharge, and coalesced the field-calibrated depth and velocity maps to create Froude and Reynolds number maps. Finally, we performed an unsupervised classification model to delineate the hydraulically relevant habitat units for salmonids. This approach provides an unprecedented view of catchment-scale hydraulic habitats that will advance both hydrological process research and river resources management.

Rapid discrimination of Salvia miltiorrhiza according to their geographical regions by laser induced breakdown spectroscopy (LIBS) and particle swarm optimization-kernel extreme learning machine (PSO-KELM)

Laser-induced breakdown spectroscopy (LIBS) coupled with particle swarm optimization-kernel extreme learning machine (PSO-KELM) method was developed for classification and identification of six types Salvia miltiorrhiza samples in different regions. The spectral data of 15 Salvia miltiorrhiza samples were collected by LIBS spectrometer. An unsupervised classification model based on principal components analysis (PCA) was employed first for the classification of Salvia miltiorrhiza in different regions. The results showed that only Salvia miltiorrhiza samples from Gansu and Sichuan Province can be easily distinguished, and the samples in other regions present a bigger challenge in classification based on PCA. A supervised classification model based on KELM was then developed for the classification of Salvia miltiorrhiza, and two methods of random forest (RF) and PSO were used as the variable selection method to eliminate useless information and improve classification ability of the KELM model. The results showed that PSO-KELM model has a better classification result with a classification accuracy of 94.87%. Comparing the results with that obtained by particle swarm optimization-least squares support vector machines (PSO-LSSVM) and PSO-RF model, the PSO-KELM model possess the best classification performance. The overall results demonstrate that LIBS technique combined with PSO-KELM method would be a promising method for classification and identification of Salvia miltiorrhiza samples in different regions.

Unsupervised classification of specialty coffees in Homogeneous sensory attributes through machine learning

Brazil is the largest exporter of coffee beans, 29% world exports, 15% this volume in specialty coffees. Thereby researches are done, so that identify different segments in the market, in order to direct the end consumer to a better quality product. New technologies are explored to meet an increasing demand for high quality coffees. Therefore, in this article has an objective to propose the use of machine learning techniques combined with projection pursuit in the construction of unsupervised classification models, in a sensory acceptance experiment, applied to four groups of trained and untrained consumers, in four classes of specialty coffees in which they were evaluated sensory characteristics: aroma, body coffee, sweetness and general note. For evaluating classifier performance, in the data with reduced dimension, all instances were used, and considering four groupings, the models were adjusted. The results obtained from the groupings formed were compared with pre-established classes to confirm the model. Success and error rates were obtained, considering the rate of false positives and false negatives, sensitivity and classification methods accuracy. It was concluded that, machine learning use in data with reduced dimensions is feasible, as it allows unsupervised classification of specialty coffees, produced at different altitudes and processes, considering the heterogeneity among consumers involved in sensory analysis, and the high homogeneity of sensory attributes among the analyzed classes, obtaining good hit rates in some classifiers. Key words: Classification models; Data dimension reduction; Groupings identification; Projection pursuit.

Unsupervised automated monitoring of dairy cows’ behavior based on Inertial Measurement Unit attached to their back

Automated monitoring of dairy cow behavior based on non-invasive sensors offers a great potential to improve the monitoring processes of its welfare and health in the context of the smart farm. It can detect any changes before the appearance of the clinical signs, allowing the farmer to take necessary measures as soon as possible. The objective of this study is to develop an effective un-supervised classification model of data collected by Inertial Measurement Units (IMU) attached to the back of dairy cows housed in free-stall. These data were aggregated according to different sampling frequencies and segmentation windows. The different times of lying, standing, lying down, standing up, walking and stationary behaviors were observed and recorded in real time. The designed classification model is based on univariate and multivariate Finite Mixture Models (FMM) and decision trees. The valid transitions between standing and lying behaviors are guaranteed by constraints imposed by a deterministic finite state automaton. The obtained results revealed that 99% of behaviors are well classified. Standing, lying on each side and changing between these positions are classified with 100% accuracy, followed by stationary with 99% sensitivity, 96% specificity, 99% precision and 99% accuracy. The walking behavior is classified with 96% sensitivity, 99% specificity, 91% precision and 98% accuracy. These results show that the back is an interesting location for sensors to monitor the dairy cow behavior.

Unsupervised Classification for Landslide Detection from Airborne Laser Scanning

Landslides are natural disasters that cause extensive environmental, infrastructure and socioeconomic damage worldwide. Since they are difficult to identify, it is imperative to evaluate innovative approaches to detect early-warning signs and assess their susceptibility, hazard and risk. The increasing availability of airborne laser-scanning data provides an opportunity for modern landslide mapping techniques to analyze topographic signature patterns of landslide, landslide-prone and landslide scarred areas over large swaths of terrain. In this study, a methodology based on several feature extractors and unsupervised classification, specifically k-means clustering and the Gaussian mixture model (GMM) were tested at the Carlyon Beach Peninsula in the state of Washington to map slide and non-slide terrain. When compared with the detailed, independently compiled landslide inventory map, the unsupervised methods correctly classify up to 87% of the terrain in the study area. These results suggest that (1) landslide scars associated with past deep-seated landslides may be identified using digital elevation models (DEMs) with unsupervised classification models; (2) feature extractors allow for individual analysis of specific topographic signatures; (3) unsupervised classification can be performed on each topographic signature using multiple number of clusters; (4) comparison of documented landslide prone regions to algorithm mapped regions show that algorithmic classification can accurately identify areas where deep-seated landslides have occurred. The conclusions of this study can be summarized by stating that unsupervised classification mapping methods and airborne light detection and ranging (LiDAR)-derived DEMs can offer important surface information that can be used as effective tools for digital terrain analysis to support landslide detection.

Using self-organizing maps for unsupervised analysis of radar data for nowcasting purposes

Predicting weather, and particularly severe weather, is an important challenge both for meteorological and machine learning researchers. The complexity and difficulty of the problem is mainly due to the chaotic character of the atmosphere and the implicit large set of meteorological information (radar, satellite or ground meteorological observations) which have to be analyzed by meteorologists. Thus, understanding the relationships between various meteorological parameters extracted from radar observations may be useful for providing additional comprehension about severe weather development and would help to identify situations when severe weather can occur. Self-organizing maps are being explored as an unsupervised classification model for detecting patterns in radar data which are relevant in predicting short-term weather changes. Experiments are performed on real radar data provided by the Romanian National Meteorological Administration. With the main goal of analyzing how the values for the weather radar products are evolving between consecutive radar scans, we empirically show that in general there is a slow change in the values over time, except for the situations when certain severe phenomena occur. The study conducted in this paper is aimed to provide a better insight regarding how the values of weather radar products are evolving in time both in calm and severe weather conditions, with the broader goal of using these findings for weather nowcasting.