Noisy Datasets Research Articles

High-fidelity simulation of FRP-confined concrete-filled steel tubes: the synergy of empirical and machine learning techniques

Abstract The present research addresses a significant gap in the current literature by overcoming the limitations associated with small, noisy datasets commonly used to predict the axial load-carrying capacity (ALC) of fiber-reinforced polymer (FRP)-encased concrete-filled steel tube compression examples (FCFST). Specifically, the authors present a refined, large-scale database that facilitates the evaluation of the prediction accuracies of three modeling techniques: finite element modeling (FEM), analytical modeling, and artificial neural networks (ANN). This comprehensive comparative analysis, underpinned by a robust experimental dataset, not only enhances predictive accuracy but also provides valuable engineering insights. Unlike previous studies, which often lack data refinement or fail to compare multiple modeling approaches, our work offers a more rigorous and holistic evaluation. The current study aims to recommend and compare the estimates of FEM, analytical model, and ANN model for capturing the ALC of FCFST examples. A database comprising 335 FCFST columns was constructed from previous studies to propose FEM and ANN models while the analytical model was proposed based on a database of 698 samples and encasing mechanics of steel tube and FRP wraps. The concrete damage plastic model was used for concrete along with bilinear and linear elastic models for steel tubes and FRP wraps, respectively. Analytical and ANN models effectively considered the lateral encasing mechanism of FCFST columns for accurate predictions. The FEM exhibited high accuracy with statistical parameters: MAE = 223.76, MAPE = 285.32, R² = 0.943, RMSE = 210.43, and a20-index = 0.83. In contrast, the ANN model outperformed, with MAE = 195, MAPE = 229.67, R² = 0.981, RMSE = 174, and a20-index = 0.89. The R² values between the models indicated strong correlations: FEM vs. analytical (0.876), analytical vs. ANN (0.914), and ANN vs. FEM (0.945), with the ANN model showing the best accuracy.

Noise-Aware Network with Shared Channel-Attention Encoder and Joint Constraint for Noisy Speech Separation

Recently, significant progress has been made in the end-to-end single-channel speech separation in clean environments. For noisy speech separation, existing research mainly uses deep neural networks to implicitly process the noise in speech signals, which does not fully utilize the impact of noise reconstruction errors on network training. We propose a lightweight noise-aware network with shared channel-attention encoder and joint constraint, named NSCJnet, which aims to improve the speech separation system performance in noisy environments. Firstly, to reduce network parameters, the model uses a parameter sharing channel attention encoder to convert noisy speech signals into a feature space. In addition, the channel attention layer (CAlayer) in encoder enhances the network's representational capacity and separation performance in noisy environments by calculating different weights of the filters in the convolution. Secondly, to make the network converge quickly, we regard noise as an estimation target of equal significance to speech, which compel the network to separate residual noise from the estimated speech, effectively suppressing lingering noise within the speech signal. Furthermore, by integrating a multi-resolution frequency constraint into the time domain loss, we introduce a weighted time-frequency joint loss constraint, empowering the network to acquire information across both dimensions to conducive to separating mixed speech with noise.It automatically strengthens important features for separation and suppresses unimportant ones during the learning process. The results on the noisy WHAM! dataset and the noisy Libri2Mix dataset show that our method has less computational complexity, and outperforms some advanced methods in various speech quality and intelligibility metrics.

An Attention‐Driven Hybrid Deep Neural Network for Enhanced Heart Disease Classification

ABSTRACTHeart disease continues to be a primary cause of mortality globally, highlighting the critical necessity for efficient early prediction and classification techniques. This study presents a new hybrid model attention‐based CNN‐Bi‐LSTM that integrates the SMOTE with an attention‐driven improved convolutional neural network‐recurrent neural network architecture to improve the classification of heart sounds, especially from imbalanced datasets. Heart sounds are difficult to classify because of their complex acoustic properties and the variability of their characteristics across frequency and temporal domains. The proposed model utilises an advanced CNN to effectively extract global and local features, in conjunction with a bidirectional long short‐term memory network to improve the architecture by capturing contextual information from both preceding and subsequent time sequences. The incorporation of spatial attention within the CNN and temporal attention in the RNN enables the model to concentrate on the most pertinent audio segments. To address the challenges presented by imbalanced and noisy datasets that may impede the efficacy of deep learning algorithms, our model employs SMOTE to improve data representation. The hybrid model outperformed popular models such as CNN, LSTM and CNN‐LSTM, achieving a classification accuracy of more than 97% on the PCG and PASCAL heart sound datasets. The findings demonstrate the model's reliability as an initial evaluation tool in clinical settings, thereby improving support for cardiovascular disease diagnosis.

Enhancing urban flow prediction via mutual reinforcement with multi-scale regional information

Intelligent Transportation Systems (ITS) are essential for modern urban development, with urban flow prediction being a key component. Accurate flow prediction optimizes routes and resource allocation, benefiting residents, businesses, and the environment. However, few methods address the spatial–temporal heterogeneity of urban flows. Existing methods typically capture spatial features solely from urban flows, but spatial feature sensitivity becomes a bottleneck when dealing with small or noisy datasets. To address this issue, we propose a method for urban flow prediction via mutual reinforcement with multi-scale regional information (MR-UFP). Firstly, we employ spatial–temporal random masking and spatial–temporal contrastive learning pre-training to directly mine spatial–temporal heterogeneity from historical flow data. Secondly, we transform the task of spatial feature extraction and embedding for urban flow prediction into a mutual reinforcement task by multi-scale region classification auxiliary task. The adaptive environment fusion module and real-time graph processing dynamically correlate regional and environmental features during flow prediction. To balance the mutual reinforcement of the two tasks, we design a joint loss function to optimize feature embedding and feedback correction, ensuring robust and accurate urban flow prediction. Extensive experiments on two real-world datasets demonstrate that MR-UFP outperforms baseline models, showcasing its robustness and effectiveness even with minimal data.

Adaptive gravitational clustering algorithm integrated with noise detection

Clustering analysis is frequently used in data mining, image processing, artificial intelligence, and so on. Traditional approaches heavily rely on manually configured parameters, of which the initial selection exerts a profound influence on the clustering outcomes. In addition, they usually only consider the relationship between two individual samples when calculating distances, neglecting the overall structure of the dataset, which can negatively affect clustering performance. At the same time, many contemporary algorithms are tailored to specific datasets, posing challenges in achieving optimal clustering performance for intricate, noisy datasets. To address these limitations, we propose an Adaptive Gravitational Clustering Algorithm Integrated with Noise Detection called GCIND. Inspired by the law of gravitation, GCIND takes into account the natural neighborhood structure of the entire dataset, adaptively computing the gravitation between data points by leveraging shared neighbors and Euclidean distance relationships. Our algorithm initially identifies and eliminates outliers or edge points in the dataset. It subsequently uses gravitation to autonomously cluster the remaining core data. Finally, the removed data are reallocated to their respective clusters. GCIND has four notable advantages: (1) it uses gravitation to build the neighborhood graph, reflecting the overall dataset structure; (2) it demonstrates stronger robustness in handling noisy datasets; (3) it uses adaptive gravitational neighborhood graph clustering, removing manual parameter tuning; (4) it adapts to complex manifold-structured datasets, offering broad applicability. Experiments have shown that GCIND, without requiring any parameter settings, demonstrates slightly better performance than the algorithms compared in the study, especially when dealing with complex manifold datasets.

Improving Differentiation of Crohn's Disease and Ulcerative Colitis Proteomes through Protein-Wide Association Study Feature Selection in Machine Learning.

Diagnostic differentiation between Crohn's disease (CD) and ulcerative colitis (UC) is crucial for timely and suitable therapeutic measures. The current gold standard for differentiating between CD and UC involves endoscopy and histology, which are invasive and costly. We aimed to identify blood plasma proteomic signatures using a Protein-Wide Association Study (PWAS) approach to differentiate CD from UC and evaluate the efficacy of these signatures as features in machine learning (ML) classifiers. Among participants (n=1,106; n CD =636; n UC =470) of the Study of a Prospective Adult Research Cohort with IBD (SPARC), plasma protein (n=2,920) levels were estimated using Olink proteomics. A PWAS with Bonferroni correction for multiple testing was used to identify proteins associated with disease states after controlling for age, sex, and disease severity. ML classifiers examined the diagnostic utility of these models. Feature importance was determined via SHapley Additive exPlanations (SHAP) analysis. Thirteen proteins which were significantly differentially abundant in CD vs UC (all |β|s > 0.22, all adjusted p values < 8.42E-06). Random forest models of proteins differentiated between CD and UC with models trained only on PWAS identified proteins (Average ROC-AUC 0.73) outperforming models trained of the full proteome (Average ROC-AUC 0.62). SHAP analysis revealed that Granzyme B, insulin-like peptide 5 (INSL5), and interleukin-12 subunit beta (IL-12B) were the most important features. Our findings demonstrate that PWAS-based feature selection approaches are a powerful method to identify features in complex, noisy datasets. Importantly, we have identified novel peptide based biomarkers such as INSL5, that can be potentially used to complement existing strategies to differentiate between CD and UC.

Research on temperature compensation of MEMS pressure sensors based on optimized multi-kernel relevance vector machine

Abstract The high-temperature coefficient in MEMS pressure sensors significantly affects sensor accuracy. Software compensation effectively reduces temperature drift. However, compensation methods such as neural networks tend to overfit sparse and noisy datasets used for calibration, leading to reduced measurement accuracy. To address the above-mentioned crucial problems, this paper proposes a multi-kernel relevance vector machine (MKRVM) based on a snow ablation optimizer (SAO) to establish the temperature-voltage-pressure relationship of pressure sensors. Optimized weight distribution from this method mitigates noise impact on the model in sparse data scenarios. Experimental results show that within the range of -15°C to 60°C, the proposed temperature compensation method achieves a mean square error as low as 0.058, effectively reducing temperature influence on pressure sensor outputs and ensuring high precision performance of the sensor.

A new binary classifier robust on noisy domains based on kNN algorithm

Classification is an effective technique commonly used in data analysis by systematically arranging groups or categories according to established criteria. The classifier's success relies on the classifier itself and the quality of the data. However, in real-world applications, it is inevitable for datasets to contain mislabeled instances, which may cause misclassification challenges that classifiers have to handle. This study aims for a quantitative assessment of the classification of noisy data through a new kNN-based classification algorithm and to increase the performance of classical kNN by efficiently classifying the data. We perform various numerical experiments on real-world data sets to prove our new algorithm's performance. We obtain high standards of accuracy levels on various noisy datasets. We propose that this new technique can provide high standard accuracy levels in binary classification problems. We compared the new kNN and classical kNN algorithms in various noise levels (10%, 20%, 30%, and 40%) on distinct datasets by measuring in terms of test accuracy. Also, we compared our new algorithm with popular classification algorithms and in the vast majority, we obtained better test accuracy results.

Learning Hamiltonian dynamics with reproducing kernel Hilbert spaces and random features

A method for learning Hamiltonian dynamics from a limited and noisy dataset is proposed. The method learns a Hamiltonian vector field on a reproducing kernel Hilbert space (RKHS) of inherently Hamiltonian vector fields, and in particular, odd Hamiltonian vector fields. This is done with a symplectic kernel, and it is shown how the kernel can be modified to an odd symplectic kernel to impose the odd symmetry. A random feature approximation is developed for the proposed odd kernel to reduce the problem size. The performance of the method is validated in simulations for three Hamiltonian systems. It is demonstrated that the use of an odd symplectic kernel improves prediction accuracy and data efficiency, and that the learned vector fields are Hamiltonian and exhibit the imposed odd symmetry characteristics.

Robust Trend Analysis in Environmental Remote Sensing: A Case Study of Cork Oak Forest Decline

We introduce a novel methodological framework for robust trend analysis (RTA) using remote sensing data to enhance the accuracy and reliability of detecting significant environmental trends. Our approach sequentially integrates the Theil–Sen (TS) slope estimator, the Contextual Mann–Kendall (CMK) test, and the false discovery rate (FDR) control. This comprehensive method addresses common challenges in trend analysis, such as handling small, noisy datasets with outliers and issues related to spatial autocorrelation, cross-correlation, and multiple testing. We applied this RTA workflow to study tree cover trends in Los Alcornocales Natural Park (Southern Spain), Europe’s largest cork oak forest, analysing interannual changes in tree cover from 2000 to 2022 using Terra MODIS MOD44B data. Our results reveal that the TS estimator provides a robust measure of trend direction and magnitude, but its effectiveness is dramatically enhanced when combined with the CMK test. This combination highlights significant trends and effectively corrects for spatial autocorrelation and cross-correlation, ensuring that genuine environmental signals are distinguished from statistical noise. Unlike previous workflows, our approach incorporates the FDR control, which successfully filtered out 29.6% of false discoveries in the case study, resulting in a more stringent assessment of true environmental trends captured by multi-temporal remotely sensed data. In the case study, we found that approximately one-third of the area exhibits significant and statistically robust declines in tree cover, with these declines being geographically clustered. Importantly, these trends correspond with relevant changes in tree cover, emphasising the ability of RTA to detect relevant environmental changes. Overall, our findings underscore the crucial importance of combining these methods, as their synergy is essential for accurately identifying and confirming robust environmental trends. The proposed RTA framework has significant implications for environmental monitoring, modelling, and management.

EEG-based emotional valence and emotion regulation classification: a data-centric and explainable approach

Emotion classification using electroencephalographic (EEG) data is a challenging task in the field of Artificial Intelligence. While many researchers have focused on finding the best model or feature extraction technique to achieve optimal results, few have attempted to select the best methodological steps for working with the dataset. In this study, we applied two different theoretical approaches based on the noise of the dataset: curriculum learning and confident learning. Curriculum learning involves presenting training examples to the model in a specific order, starting with easier examples and gradually increasing in difficulty. This approach has been shown to improve model performance. Confident learning is a method for identifying and correcting label errors in datasets. By identifying and correcting these errors, confident learning can improve the performance of machine learning models trained on noisy datasets. We then applied the Integrated Gradient technique in order to assess the explainability of each model. Our aim was to explore the impact of different models and methods on emotion classification performance using EEG data. We collected and used an EEG dataset in which participants rated the emotional valence of positive and negative pictures while performing an emotion regulation (ER) task, comparing a control condition (Look) with two ER strategies: cognitive reappraisal and expressive suppression. We performed a multilabel classification to identify emotional neutrality or polarization of emotional valence (both positive and negative) rated by participants and the emotion regulation strategy adopted during the task. We compared the performance of models trained on three datasets selected based on label noise and evaluated their suitability for this task. Our results suggest different patterns based on the architecture used for feature importance, highlighting both advantages and criticisms.

A deep autoencoder for electric double layer capacitance prediction in electrochemical sensors

This study explores the application of a deep autoencoder neural network to accurately predict the electric double layer capacitance from real-world parameters in binary, asymmetric electrolytes under low concentration conditions. By utilizing a modest simulation-based dataset of just 250 samples, the deep autoencoder neural network model developed in this study effectively predicted the capacitance by learning the critical features and relationships of the electric double layer model and encoding this learned representation into a low-dimensional latent space. From the latent variables, the decoder block of the neural network learned to effectively recreate the high-dimensional input. To enhance the model's robustness, prevent overfitting, and better simulate real-world conditions, noise was incorporated into the training and test data. The model demonstrated strong performance across various conditions, such as ionic size, ionic charge, and surface potential, yielding satisfactory results on both clean and noisy test datasets. A key feature of this approach was the mapping of real-world electric double layer parameters to the latent variables of the model, allowing for direct input of physical parameters to predict the electric double layer capacitance. This research highlights the potential of machine learning techniques to expedite the design and analysis of complex multi-physics systems such as electrochemical sensors by reducing the dependence on extensive domain expertise throughout the design process.

Uncertainty in Fourier Transforms: A Fuzzy Logic Perspective

This study explores the integration of fuzzy logic with Fourier transforms to address the challenges of uncertainty, noise, and imprecision in real-world data. Fuzzy Fourier transforms extend traditional Fourier methods by incorporating fuzzy numbers, allowing for more robust frequency analysis, signal processing, and data reconstruction, particularly in noisy or incomplete datasets. The study examines the mathematical formulation of fuzzy Fourier transforms, computational trade-offs, and their performance advantages in comparison to classical Fourier methods. Real-world applications are discussed, including signal processing, image reconstruction, and time-series forecasting. Furthermore, the research highlights the increased computational complexity associated with fuzzy methods, the challenges of interpreting fuzzy results, and the limitations of handling probabilistic uncertainty. Future directions include the development of multi-dimensional fuzzy Fourier transforms, hybrid models integrating machine learning, and quantum computing applications. These advancements have broad implications for fields such as telecommunications, medical imaging, and financial forecasting, where handling uncertainty is critical.

MTC-NET: A Multi-Channel Independent Anomaly Detection Method for Network Traffic.

In recent years, deep learning-based approaches, particularly those leveraging the Transformer architecture, have garnered widespread attention for network traffic anomaly detection. However, when dealing with noisy data sets, directly inputting network traffic sequences into Transformer networks often significantly degrades detection performance due to interference and noise across dimensions. In this paper, we propose a novel multi-channel network traffic anomaly detection model, MTC-Net, which reduces computational complexity and enhances the model's ability to capture long-distance dependencies. This is achieved by decomposing network traffic sequences into multiple unidimensional time sequences and introducing a patch-based strategy that enables each sub-sequence to retain local semantic information. A backbone network combining Transformer and CNN is employed to capture complex patterns, with information from all channels being fused at the final classification header in order to achieve modelling and detection of complex network traffic patterns. The experimental results demonstrate that MTC-Net outperforms existing state-of-the-art methods in several evaluation metrics, including accuracy, precision, recall, and F1 score, on four publicly available data sets: KDD Cup 99, NSL-KDD, UNSW-NB15, and CIC-IDS2017.

Segmenting hotspots from medical thermal images using Density-based modified FC-Pc FS with spatial information

ABSTRACT Thermal cameras complemented with robust and automated computational intelligence-aided diagnostic systems can be deployed at public gathering points and on smartphones to raise alarms about thriving diseases of their body due to inflammation in a timely manner, thereby strengthening human health systems. In this work, we have proposed two robust Density-based modified Picture Fuzzy Clustering methods with spatial information to segment hotspots from human’s abnormal thermal images. We formulated and solved the objective function by combining P c FS-based clustering method with modified Renyi’s Entropy. We demonstrated the robustness of our methods, DSIFC-Pc FS and DSIMFC-Pc FS, on three publicly available datasets and an artificially created noisy dataset of thermal images. We have obtained a ZC score of 0.933, 0.947 and 0.933 for DB-DMR-IR, DB-FOOT-IR and DB-THY-IR datasets, respectively. Also, our method converged to an optimal partition matrix in minimum iterations. The experimental performance shows that the proposed approaches, DSIFC- P c FS and DSIMFC- P c FS, outperformed other ten related approaches in segmenting the medical thermal images with and without noise, indicating their robustness and efficacy, in terms of various unsupervised evaluation metrics. Also, we statistically analysed their performance in contrast to other methods using the Friedman Test, which advocate their superiority.

Pose‐to‐Motion: Cross‐Domain Motion Retargeting with Pose Prior

AbstractCreating plausible motions for a diverse range of characters is a long‐standing goal in computer graphics. Current learning‐based motion synthesis methods rely on large‐scale motion datasets, which are often difficult if not impossible to acquire. On the other hand, pose data is more accessible, since static posed characters are easier to create and can even be extracted from images using recent advancements in computer vision. In this paper, we tap into this alternative data source and introduce a neural motion synthesis approach through retargeting, which generates plausible motion of various characters that only have pose data by transferring motion from one single existing motion capture dataset of another drastically different characters. Our experiments show that our method effectively combines the motion features of the source character with the pose features of the target character, and performs robustly with small or noisy pose data sets, ranging from a few artist‐created poses to noisy poses estimated directly from images. Additionally, a conducted user study indicated that a majority of participants found our retargeted motion to be more enjoyable to watch, more lifelike in appearance, and exhibiting fewer artifacts. Our code and dataset can be accessed here.

MD-BiMamba: An aero-engine inter-shaft bearing fault diagnosis method based on Mamba with modal decomposition and bidirectional features fusion strategy

Inter-shaft bearing fault diagnosis can greatly save maintenance costs and guarantee the normal operation of aero-engines. However, existing methods are limited in their ability to handle sensor signals with high dimensional and complex noise. To tackle the above problems, we propose a Mamba network structure with Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) and bidirectional feature fusion for aero-engine inter-shaft bearing fault diagnosis, called MD-BiMamba. Firstly, we utilize the CEEMDAN technique to decompose the sensor signals of inter-shaft bearings and extract the intrinsic modal function by improving the noise addition method and decomposition strategy. Then, bidirectional Mamba is designed to extract the bidirectional long-range correlation of the sequence to achieve global feature extraction and fast detection of fault types with linear complexity. Moreover, an adaptive attention fusion method is proposed to achieve bidirectional feature fusion. Finally, we use the label smoothing regularization strategy to optimize the cross-entropy loss and enhance the generalization performance. Experimental results on real inter-shaft bearing datasets and noisy datasets show that the proposed method achieves 99.88 % and 95.81 % accuracy in the standard dataset and −10 dB noise environment, respectively, which is superior to other remarkable methods, in addition to achieving lower model complexity.

NEURO DEEP FUZZY GENETIC ALGORITHM APPROACH FOR CLASSIFICATION AND DETECTION OF BRAIN TUMOR FROM LARGE DATASETS

Brain tumors represent one of the most critical and life-threatening diseases. Early detection and accurate classification are essential for effective treatment planning and survival rate improvement. With the exponential growth of medical data, particularly in imaging and diagnostic datasets, traditional algorithms face limitations in handling large-scale, high-dimensional data efficiently. Conventional machine learning and diagnostic methods often struggle with classification accuracy, computational complexity, and overfitting in large, noisy datasets. Addressing these issues is crucial for the development of robust diagnostic tools capable of handling real-world clinical data. This paper presents a novel hybrid approach combining Neural Networks, Deep Learning, Fuzzy Logic, and Genetic Algorithms, termed Neuro Deep Fuzzy Genetic Algorithm (NDFGA), for the classification and detection of brain tumors from large datasets. Neural Networks and Deep Learning architectures are leveraged for feature extraction and hierarchical learning. Fuzzy Logic improves interpretability and manages uncertainty in medical data, while Genetic Algorithms optimize feature selection and model parameters. This hybrid method is designed to maximize classification accuracy while minimizing false positives and computational overhead. The proposed approach was tested on a large brain tumor dataset comprising over 10,000 MRI scans. The NDFGA approach demonstrated superior performance compared to standalone methods. It achieved a classification accuracy of 97.8%, a sensitivity of 96.5%, and a specificity of 98.1%. The model also showed improved robustness in handling large datasets, reducing false positives by 12% and computational time by 15% compared to traditional methods. This hybrid model presents a scalable and efficient solution for brain tumor detection, especially in clinical environments.

Systematization of sunflower genotypes based on seed phenotypic characteristics using neural networks

In recent years, various datasets related to the phenotyping of sunflower genotypes have become increasingly accessible. However, one of the key challenges remains the efficient and accurate prediction of phenotypes based on genotypes in the context of climate change. Analyzing phenotypes at different levels of organization and detecting connections between phenotypes and genotypes require the integration and processing of large, diverse, and often noisy datasets. Machine learning offers a broad arsenal of methods and approaches for identifying predictive patterns in such data. Therefore, the research aimed to develop a methodology for the systematization of sunflower genotypes based on seed phenotypic characteristics using the data vector quantization method and neural networks. The study revealed the phenotypic characteristics of sunflower seeds from various genotypes selected by the Institute of Oilseed Crops of NAAS, grown in the southern Steppe of Ukraine, including seed length, width, thickness, seed mass, kernel mass, and seed coat cracking force. For this purpose, appropriate laboratory equipment was developed, including two modules for determining the morphological and rheological properties of seeds. The developed methodology for the systematization of sunflower genotypes based on seed phenotypic characteristics includes the following steps: measuring the characteristics of sunflower seeds from various samples (parental components); studying the mutual correlation of characteristics; conducting hierarchical cluster analysis of the data using the Ward's method; determining the optimal number of groups; performing k-means clustering using the vector quantization method; determining the correspondence of ranges of characteristics to the group; training a neural network to group the data by samples and created groups; verifying the adequacy of the neural network on test data. The developed methodology was tested, and the MLP 30-15-3 neural network for grouping data by samples and created groups of sunflower seeds was developed in the Statistica software package. The network's training efficiency was 99.4%, and such of testing and validation was 95.6% and 96.7%, respectively.

Artificial-Intelligence-Based Condition Monitoring of Industrial Collaborative Robots: Detecting Anomalies and Adapting to Trajectory Changes

The increasing use of collaborative robots in smart manufacturing, owing to their flexibility and safety benefits, underscores a critical need for robust predictive maintenance strategies to prevent unexpected faults/failures of the machine. This paper focuses on fault detection and employs multivariate operational data from a universal robot to detect anomalies or early-stage faults using test data from designed anomalous conditions and artificial-intelligence-based anomaly detection techniques called autoencoders. The performance of three autoencoders, namely, a multi-layer-perceptron-based autoencoder, convolutional-neural-network-based autoencoder, and sparse autoencoder, was compared in detecting anomalies. The results indicate that the autoencoders effectively detected anomalies in the examined complex and noisy datasets with more than 93% overall accuracy and an F1 score exceeding 96% for the considered anomalous cases. Moreover, the integration of trajectory change detection and anomaly detection algorithms (i.e., the dynamic time warping algorithm and sparse autoencoder, respectively) was proposed for the local implementation of online condition monitoring. This integrated approach to anomaly detection and trajectory change provides a practical, adaptive, and economical solution for enhancing the reliability and safety of collaborative robots in smart manufacturing environments.