Action recognition has witnessed the development of a growing number of novel algorithms and datasets in the past decade. However, the majority of public benchmarks were constructed around activities of daily living and annotated at a rather coarse-grained level, which lacks diversity in domain- specific datasets, especially for rarely seen domains. In this paper, we introduced Human Stone Toolmaking Action Grammar (HSTAG), a meticulously annotated video dataset showcasing previously undocumented stone toolmaking behaviors, which can be used for investigating the applications of advanced artificial intelligence techniques in understanding a rapid succession of complex interactions between two hand-held objects. HSTAG consists of 18,739 video clips that record 4.5 hours of experts' activities in stone toolmaking. Its unique features include (i) brief action durations and frequent transitions, mirroring the rapid changes inherent in many motor behaviors; (ii) multiple angles of view and switches among multiple tools, increasing intra-class variability; (iii) unbalanced class distributions and high similarity among different action sequences, adding difficulty in capturing distinct patterns for each action. Several mainstream action recognition models are used to conduct experimental analysis, which showcases the challenges and uniqueness of HSTAG.

This paper proposes a novel data-driven technique, namely Online Learning for sensor Fault diagnosis Analysis (OLFA), to perform real-time fault analysis for autonomous cars. Considering the non-stationary properties of real-time sensor faults and the mapping relationship between sensors and feature variables, the proposed method decomposes the sensor fault diagnosis analysis problem into an online data stream classification and feature ranking problems. To detect and identify faults, a clustering-based data stream classification approach is developed to continuously capture and classify non-stationary sensor faults for autonomous cars with little intervention from human experts. An effective active learning method is extended and embedded into the proposed framework to minimize the need for prior knowledge about faults and enable the continual learning capability to adapt to and handle the non-stationary properties of sensor faults. Moreover, the proposed framework addresses the parameter optimization issue of existing machine learning based fault analysis techniques and employs feature ranking analysis to systematically analyze the possible source(s) of sensor faults. CAR Learning to Act (CARLA), a well-known realistic autonomous driving simulator, is used as the benchmark to perform the sensor fault injection and online data stream collection to evaluate the efficacy of OLFA. Analysis of the collected faulty datasets and experimental results, and comparison between OLFA and several state-of-the-art clustering-based approaches for fault classification, demonstrated the efficacy of the proposed framework in the domain of autonomous cars.

Increased traffic density with a greater degree of increased automation in aviation is expected within the next decade. Therefore, airspace capacity will become more congested and result in increasing challenges for detecting conflicts between aerial vehicles. Furthermore, because these vehicles rely on surrounding vehicles following a planned path, it is essential to identify flights not following a planned direction. In this paper, we utilize an ensemble of the existing outlier detection approaches for identifying the anomalous flight trajectories. In the initial step, flight trajectories are preprocessed to extract and process vital features, with the next step of having twenty different outlier detection algorithms assembled to classify trajectories. Throughout our extensive experiments and comparison studies, promising results are shown including the effectiveness of different anomaly detection algorithms and how utilizing feature engineering can improve the results of these outlier detection methods.

Realtime estimation of travel time is a key traffic parameter for designing and planning for transportation systems, particularly when providing mobility-on-demand (MOD) services. However, the analysis and prediction of travel time can be delayed significantly due to the complexity and huge computational requirements of microsimulation models. Thus, as an alternative solution, we propose a data-driven approach for the efficient and reliable prediction of travel time. Our approach takes advantage of the strengths of SVM and ARIMA for fully capturing the traffic patterns in the traffic data. We introduce a new parameter $\kappa$ into the SVM-ARIMA model to adjust the weight of the ARIMA component, which significantly improves the performance. We validate the performance of the proposed approach using data generated from a microsimulation platform. Our experimental results and comparisons with the existing ML-based methods demonstrates the efficacy of the proposed data-driven approach.

The purpose of this paper is to devise an interpretable hybrid classification model for Convolutional Neural Networks (CNN) and a Learning Classifier System (LCS). The presented hybrid system integrates the fundamental attributes from both types of these classifiers. In the proposed hybrid model CNN works as an automatic feature extractor, and LCS works to provide interpretable rule-based classification results. Although LCS has limitations working on higher dimensional datasets, we resolve this limitation by using CNN as a feature extractor. The other concept of the non-interpretability of CNN is addressed by using the LCS rule. Furthermore, our experiment with higher dimensional datasets like CIFAR-10 and Fashion-MNIST shows that extended LCS provides comparable performance to the standard neural network model while also providing interpretable results. We named this extended LCS method Convolutional Learning Classifier Cystem (C-LCS).

The non-stationary nature of data streams strongly challenges traditional machine learning techniques. Although some solutions have been proposed to extend traditional machine learning techniques for handling data streams, these approaches either require an initial label set or rely on specialized design parameters. The overlap among classes and the labeling of data streams constitute other major challenges for classifying data streams. In this paper, we proposed a clustering-based data stream classification framework to handle non-stationary data streams without utilizing an initial label set. A density-based stream clustering procedure is used to capture novel concepts with a dynamic threshold and an effective active label querying strategy is introduced to continuously learn the new concepts from the data streams. The sub-cluster structure of each cluster is explored to handle the overlap among classes. Experimental results and quantitative comparison studies reveal that the proposed method provides statistically better or comparable performance than the existing methods.

Standard Convolutional Neural Network (CNN) designs rarely focus on the importance of explicitly capturing diverse features to enhance the network’s performance. Instead, most existing methods follow an indirect approach of increasing or tuning the networks’ depth and width, which in many cases significantly increase the computational cost. Inspired by biological visual system, we proposes a Diverse and Adaptive Attention Convolutional Network (DA2-Net), which enables any feed-forward CNNs to explicitly capture diverse features and adaptively select and emphasize the most informative features to efficiently boost the network’s performance. DA2-Net incurs negligible computational overhead and it is designed to be easily integrated with any CNN architecture. We extensively evaluated DA2-Net on benchmark datasets, including CIFAR100, SVHN, and ImageNet, with various CNN architectures. The experimental results show DA2-Net provides a significant performance improvement with very minimal computational overhead.

Feature selection methods are widely used to address the high computational overheads and curse of dimensionality in classifying high-dimensional data. Most conventional feature selection methods focus on handling homogeneous features, while real-world datasets usually have a mixture of continuous and discrete features. Some recent mixed-type feature selection studies only select features with high relevance to class labels and ignore the redundancy among features. The determination of an appropriate feature subset is also a challenge. In this paper, a supervised feature selection method using density-based feature clustering (SFSDFC) is proposed to obtain an appropriate final feature subset for mixed-type data. SFSDFC decomposes the feature space into a set of disjoint feature clusters using a novel density-based clustering method. Then, an effective feature selection strategy is employed to obtain a subset of important features with minimal redundancy from those feature clusters. Extensive experiments as well as comparison studies with five state-of-the-art methods are conducted on SFSDFC using thirteen real-world benchmark datasets and results justify the efficacy of the SFSDFC method.

As a promising control method of permanent magnet synchronous motors (PMSM), the finite control set model predictive control (FCS-MPC) can overcome the integral saturation and bandwidth limitation of proportional-integral controllers. However, the weight design of the cost function is indeed a serious subject, which directly affects the control performance of FCS-MPC and should be carefully considered. Currently, the weight design lacks a theoretical basis and has a strong randomness. In this paper, a novel weight design of FCS-MPC based on frequency domain response characteristics is proposed for PMSM driving systems. Different from the conventional experience-based weight design, this study extracts the frequency domain response characteristics of FCS-MPC, analyzes the complex relationships between the weights in the cost function and frequency domain response characteristics, and finally obtains new weight design rules. Some numerical analysis and simulation are carried on to prove the correctness and feasibility of the proposed weight design method.

Despite advancements in vehicle security systems, over the last decade, auto-theft rates have increased, and cyber-security attacks on internet-connected and autonomous vehicles are becoming a new threat. In this paper, a deep learning model is proposed, which can identify drivers from their driving behaviors based on vehicle telematics data. The proposed Long-Short-Term-Memory (LSTM) model predicts the identity of the driver based on the individual's unique driving patterns learned from the vehicle telematics data. Given the telematics is time-series data, the problem is formulated as a time series prediction task to exploit the embedded sequential information. The performance of the proposed approach is evaluated on three naturalistic driving datasets, which gives high accuracy prediction results. The robustness of the model on noisy and anomalous data that is usually caused by sensor defects or environmental factors is also investigated. Results show that the proposed model prediction accuracy remains satisfactory and outperforms the other approaches despite the extent of anomalies and noise-induced in the data.

Multi-sensor data fusion is widely used in various application domains. Integration of multiple sensors is a complex problem. This is because it is often characterized by uncertainty due to randomness and non-specificity. The Dempster Shafer (DS) theory of evidence has often been used for modelling and reasoning under uncertainty. However, the DS rule of combination is often prone to counter-intuitive results when combining pieces of evidence that are highly conflicting. As a result, several alternative combination rules have emerged. One approach is to assign weight to each basic probability assignment (BPA) prior to the use of the DS rule of combination. Most existing methods of assigning weight only focus on the credibility of each BPA without considering the reliability of the source of the BPA. In this work, we propose a multi-sensor data fusion that takes into consideration both the reliability of each BPA source and the credibility degree. A benchmark dataset was used to evaluate the effectiveness of the proposed method. To further assess the robustness of the proposed method in handling uncertainty, different noise levels were introduced to the training set.

Multi-label classification deals with problem domains in which each instance belongs to more than one class simultaneously. Label Powerset (LP) is an efficient multi-label learning algorithm that considers each distinct combination of labels in training data as a unique new class and trains a conventional multi-class learning algorithm. In this paper a Multi-label classification algorithm is proposed that integrates LP with a rule-based evolutionary machine learning approach developed for supervised learning tasks, namely sUpervised Learning Classifiers (UCS). Moreover, to improve the prediction capability of the model on unseen instances, a prediction aggregation strategy is proposed to make efficient use of all the potentially helpful information in the rule base. The result is a multi-label rule-based evolutionary learner, which is called MLRBC (Multi-Label Rule-Based Classifier). Taking advantage of the strong generalization capability of UCS and its robustness in handling data sets with imbalanced classes, the proposed MLRBC algorithm is able to address some of the challenges involved in using LP. Experimental studies on multiple real-world datasets show that the proposed algorithm substantially improves the performance of the original LP technique and shows competitive performance against some of the state of the art multi-label learning algorithms.

As an effective dimensionality reduction technique, feature selection is widely used in the preprocessing procedure in data mining. It is highly advocated by its superiority in mitigating the effect of noisy data and simplifying the analysis of high-dimensional data. In this paper, a novel unsupervised feature selection procedure based on a clustering algorithm is proposed to evaluate the goodness of features and select a set of useful features without losing the characteristics of the data. It consists of two steps: clustering and feature evaluation. In the clustering procedure, a novel clustering algorithm based on the fitness proportionate sharing is adopted to separate data into distinct clusters without any prior knowledge about data, which is more applicable to the analysis of unknown datasets. On the other hand, the feature evaluation procedure will use the information extracted from the clustering procedure to evaluate the usefulness of each feature and select good features. The proposed method is simulated with four other famous existing feature selection algorithms and a comparison is provided in this paper. Simulation results on both synthetic and real datasets demonstrate that the proposed procedure of feature selection can effectively evaluate the significance of features and obtain a better subset of features than other four existing algorithms.

Existing clustering techniques primarily rely on prior knowledge about the data, such as the number of clusters and radii. However, in real applications, the number of clusters and the radii of clusters are usually unknown. Therefore, the performance of clustering methods with overlapping data is degraded due to their limitations in finding all cluster centers with uneven density values. Hence, a new clustering algorithm based on fitness proportionate sharing is proposed to map the problem into a multimodal optimization problem. In this paper, clusters are considered as niches, and the individuals with the highest density values of each niche are the cluster centers. Instead of using the traditional sharing strategy, the fitness proportionate sharing strategy is implemented in the identification of niche maxima to overcome the sensitivity of uneven density values of cluster centers. A procedure of niche expansion is employed for the merging of clusters. Simulation results and complexity analysis reveal that the proposed clustering algorithm based on fitness proportionate sharing provides a higher accuracy performance without any prior information.