Search results for "550"

The Use of Cross-Platform Frameworks for Google Play Store Apps

2022

In this paper, we describe the harnessing and analyses of a large sample (n = 661705) of Android apps and associated metadata available on the Google Play Store. The analyses and scrutiny are in the context of cross-platform mobile development, as we report on the technologies used to develop apps for the Android ecosystem. Specifically, we quantify the use of 13 technical frameworks for cross-platform development, identify their distribution across Google Play Store categories, present an overview of framework usage from 2008 to 2019, app file size (.apk size), and lastly discuss our findings in the context of current industry trends and directions. Our findings indicate that cross-platfor…

Online Machine Learning for Graph Topology Identification from Multiple Time Series

2020

High dimensional time series data are observed in many complex systems. In networked data, some of the time series are influenced by other time series. Identifying these relations encoded in a graph structure or topology among the time series is of paramount interest in certain applications since the identified structure can provide insights about the underlying system and can assist in inference tasks. In practice, the underlying topology is usually sparse, that is, not all the participating time series in influence each other. The goal of this dissertation pertains to study the problem of sparse topology identification under various settings. Topology identification from time series is a …

The Relationship Between Outbound and Inbound Communication in Government-to-Citizen Interaction

2020

Synthetic Micro-Doppler Signatures of Non-Stationary Channels for the Design of Human Activity Recognition Systems

2021

The main aim of this dissertation is to generate synthetic micro-Doppler signatures and TV-MDSs to train the HACs. This is achieved by developing non-stationary fixed-tofixed (F2F) indoor channel models. Such models provide an in-depth understanding of the channel parameters that influence the micro-Doppler signatures and TV-MDSs. Hence, the proposed non-stationary channel models help to generate the micro-Doppler signatures and the TV-MDSs, which fit those of the collected measurement data. First, we start with a simple two-dimensional (2D) non-stationary F2F channel model with fixed and moving scatterers. Such a model assumes that the moving scatterers are moving in 2D geometry with simpl…

Hierarchical Object Detection applied to Fish Species

2022

Gathering information of aquatic life is often based on timeconsuming methods utilizing video feeds. It would be beneficial to capture more information cost-effectively from video feeds. Video based object detection has an ability to achieve this. Recent research has shown promising results with the use of YOLO for object detection of fish. As underwater conditions can be difficult and thus fish species are hard to discriminate. This study proposes a hierarchical structure-based YOLO Fish algorithm in both the classification and the dataset to gain valuable information. With the use of hierarchical classification and other techniques. YOLO Fish is a state-of-the-art object detector on Nordi…

Robust Interpretable Text Classification against Spurious Correlations Using AND-rules with Negation

2022

The state-of-the-art natural language processing models have raised the bar for excellent performance on a variety of tasks in recent years. However, concerns are rising over their primitive sensitivity to distribution biases that reside in the training and testing data. This issue hugely impacts the performance of the models when exposed to out-of-distribution and counterfactual data. The root cause seems to be that many machine learning models are prone to learn the shortcuts, modelling simple correlations rather than more fundamental and general relationships. As a result, such text classifiers tend to perform poorly when a human makes minor modifications to the data, which raises questi…

Affinity-Based Reinforcement Learning : A New Paradigm for Agent Interpretability

2023

The steady increase in complexity of reinforcement learning (RL) algorithms is accompanied by a corresponding increase in opacity that obfuscates insights into their devised strategies. Methods in explainable artificial intelligence seek to mitigate this opacity by either creating transparent algorithms or extracting explanations post hoc. A third category exists that allows the developer to affect what agents learn: constrained RL has been used in safety-critical applications and prohibits agents from visiting certain states; preference-based RL agents have been used in robotics applications and learn state-action preferences instead of traditional reward functions. We propose a new affini…

Hand Gesture Classification Using Grayscale Thermal Images and Convolutional Neural Network

2021

Accepted manuscript.

On abstraction in the OMG hierarchy: systems, models, and descriptions

2022

The Model-Driven Architecture (MDA) uses a metadata hierarchy with several layers that are placed on top of each other. The traditional view is that the layers provide abstractions related to models in languages defined by meta-models. Over the years, it has been difficult to define a consistent understanding of the layers. In this paper, we propose such a consistent understanding by clarifying the relations between the different elements in the hierarchy. This is done based on the Scandinavian approach to modelling that distinguishes between systems and system descriptions. Systems can be physical, digital, or even mental, while descriptions can be programs, language descriptions, specific…

LONG HORIZON ANOMALY PREDICTION IN MULTIVARIATE TIME SERIES WITH CAUSAL AUTOENCODERS

2022

Predictive maintenance is essential for complex industrial systems to foresee anomalies before major system faults or ultimate breakdown. However, the existing efforts on Industry 4.0 predictive monitoring are directed at semi-supervised anomaly detection with limited robustness for large systems, which are often accompanied by uncleaned and unlabeled data. We address the challenge of predicting anomalies through data-driven end-to-end deep learning models using early warning symptoms on multivariate time series sensor data. We introduce AnoP, a long multi-timestep anomaly prediction system based on unsupervised attention-based causal residual networks, to raise alerts for anomaly preventio…