Applying Evidence-Centered Design to Measure Psychological Resilience: The Development and Preliminary Validation of a Novel Simulation-Based Assessment Methodology

Modern technologies have enabled the development of dynamic game- and simulation-based assessments to measure psychological constructs. This has highlighted their potential for supplementing other assessment modalities, such as self-report. This study describes the development, design, and preliminary validation of a simulation-based assessment methodology to measure psychological resilience—an important construct for multiple life domains. The design was guided by theories of resilience, and principles of evidence-centered design and stealth assessment. The system analyzed log files from a simulated task to derive individual trajectories in response to stressors. Using slope analyses, these trajectories were indicative of four types of responses to stressors: thriving, recovery, surviving, and succumbing. Using Machine Learning, the trajectories were predictive of self-reported resilience (Connor-Davidson Resilience Scale) with high accuracy, supporting construct validity of the simulation-based assessment. These findings add to the growing evidence supporting the utility of gamified assessment of psychological constructs. Importantly, these findings address theoretical debates about the construct of resilience, adding to its theory, supporting the combination of the “trait” and “process” approaches to its operationalization.

Anomaly Detection using system logs

Anomaly detection is a very important step in building a secure and trustworthy system. Manually it is daunting to analyze and detect failures and anomalies. In this paper, we proposed an approach that leverages the pattern matching capabilities of Convolution Neural Network (CNN) for anomaly detection in system logs. Features from log files are extracted using a windowing technique. Based on this feature, a one-dimensional image (1×n dimension) is generated where the pixel values of an image correlate with the features of the logs. On these images, the 1D Convolution operation is applied followed by max pooling. Followed by Convolution layers, a multi-layer feed-forward neural network is used as a classifier that learns to classify the logs as normal or abnormal from the representation created by the convolution layers. The model learns the variation in log pattern for normal and abnormal behavior. The proposed approach achieved improved accuracy compared to existing approaches for anomaly detection in Hadoop Distributed File System (HDFS) logs.

Investigating Epistemic Stances in Game Play Through Learning Analytics

The authors apply techniques of statistical computing to data logs to investigate the patterns in students' play of The Fuzzy Chronicles and how these patterns relate to learning outcomes related to Newtonian kinematics. This chapter has two goals. The first goal is to investigate the basic claims of the proposed two-system framework for game-based learning (or 2SM) that may serve as part of a general-use explanatory framework for educational gaming. The second goal is to explore and demonstrate the use of automated log files of student play as evidence of learning through educational data mining techniques. These goals were pursued via two research questions. The first research question examines whether students playing the game showed evidence of dichotomous fast/slow modes of solution. A second research question investigates the connection between conceptual understanding and student performance in conceptually-laden challenges. Implications in terms of game design, learning analytics, and refinement of the 2SM are discussed.

Informing Cyber Threat Intelligence through Dark Web Situational Awareness: The AZSecure Hacker Assets Portal

To increase situational awareness, major cybersecurity platforms offer Cyber Threat Intelligence (CTI) about emerging cyber threats, key threat actors, and their modus operandi. However, this intelligence is often reactive, as it analyzes event log files after attacks have already occurred, lacking more active scrutiny of potential threats brewing in cyberspace before an attack has occurred. One intelligence source receiving significant attention is the Dark Web, where significant quantities of malicious hacking tools and other cyber assets are hosted. We present the AZSecure Hacker Assets Portal (HAP). The Dark Web-based HAP collects, analyzes, and reports on the major Dark Web data sources to offer unique perspective of hackers, their cybercriminal assets, and their intentions and motivations, ultimately contributing CTI insights to improve situational awareness. HAP currently supports 200+ users internationally from academic institutions such as UT San Antonio and National Taiwan University, law enforcement entities such as Calgary and Ontario Provincial Police, and industry organizations including General Electric and PayPal.

IoT Based Surveillance System Using DNN Model

Localization, Visibility, Proximity, Detection, Recognition has always been a challenge for surveillance system. These challenges can be felt in the industries where surveillance systems are used like armed forces, technical-agriculture and other such fields. Most of the Smart system available are just for the surveillance of Human intervention but there is a need for a system which can be used for animals as well because with the outburst of human population and symbiotic relationship with wild animals results in life loss and damage to agriculture. In this paper we are designing to overcome these above-mentioned challenges for human and animal-based surveillance system in real time application. The system setup is done on a Raspberry pi integrated with deep-learning models which performs the classification of objects on the frames, then the classified objects is given to a face detection model for further processing. The detected face is relayed to the back-end for feature mapping with the saved log files with containing features of familiar face IDs. Four models were tested for face detection out of which the DNN model performed the best giving an accuracy of 94.88%.The system is also able to send alerts to the admin if any threat is detected with the help of a communication module. Keywords: Deep learning, Raspberry Pi, OpenCV, Image Processing, YOLO, Face Recognition

Classification and Prediction of Software Incidents Using Machine Learning Techniques

An incident, in the perception of information technology, is an event that is not part of a normal process and disrupts operational procedure. This research work particularly focuses on software failure incidents. In any operational environment, software failure can put the quality and performance of services at risk. Many efforts are made to overcome this incident of software failure and to restore normal service as soon as possible. The main contribution of this study is software failure incidents classification and prediction using machine learning. In this study, an active learning approach is used to selectively label those data which is considered to be more informative to build models. Firstly, the sample with the highest randomness (entropy) is selected for labeling. Secondly, to classify the labeled observation into either failure or no failure classes, a binary classifier is used that predicts the target class label as failure or not. For classification, Support Vector Machine is used as a main classifier to classify the data. We derived our prediction models from the failure log files collected from the ECLIPSE software repository.

Building a precise machine-specific time structure of the spot and energy delivery model for a cyclotron-based proton therapy system

Objective: We proposed an experimental approach to build a precise machine-specific beam delivery time (BDT) prediction and delivery sequence model for standard, volumetric, and layer repainting delivery based on a cyclotron accelerator system. Approach Test fields and clinical treatment plans’ log files were used to experimentally derive three main beam delivery parameters that impacted BDT: energy layer switching time (ELST), spot switching time (SSWT), and spot drill time (SDT). This derived machine-specific model includes standard, volumetric, and layer repainting delivery sequences. A total of 103 clinical treatment fields were used to validate the model. Main results: The study found that ELST is not stochastic in this specific machine. Instead, it is actually the data transmission time or energy selection time, whichever takes longer. The validation showed that the accuracy of each component of the BDT matches well between machine log files and the model’s prediction. The average total BDT was about (-0.74±3.33)% difference compared to the actual treatment log files, which is improved from the current commercial proton therapy system’s prediction (67.22%±26.19%). Significance: An accurate BDT prediction and delivery sequence model was established for an cyclotron-based proton therapy system IBA ProteusPLUS®. Most institutions could adopt this method to build a machine-specific model for their own proton system.

Evaluation of the Immediate Effects of Digital Intervention Modules for Goals, Planning and Coping Planning on Physical Activity: Secondary Analysis of a Randomized Controlled Trial (Preprint)

BACKGROUND The use of digital interventions can be accurately monitored via log files. However, monitoring engagement with intervention goals or enactment of the actual behaviors targeted by the intervention is more difficult and is usually evaluated based on pre-post measurements in a controlled trial. OBJECTIVE To evaluate if engaging with two digital intervention modules focusing on physical activity goals and action plans, and coping with barriers had immediate effects on the actual physical activity behavior. METHODS The NoHoW Toolkit (TK) was a digital intervention developed for supporting long-term weight loss maintenance, evaluated in a 2 x 2 factorial randomized controlled trial. The TK contained various modules based on behavioral self-regulation and motivation theories, and contextual emotion regulation approaches, and involved continuous tracking of weight and physical activity through connected commercial devices (Fitbit Aria TM and Charge 2 TM). Two of the four trial arms had access to two modules directly targeting physical activity, i.e. a module for goal setting and action planning (“Goal”) and a module for identifying barriers and coping planning (“Barriers”). Module visits and completion were determined based on TK log files and time spent in the module web page. Five physical activity metrics (steps; activity; energy expenditure; fairly active, very active and total active minutes; and distance) were compared before and after visiting and completing the modules to examine whether the modules had immediate or sustained effects on physical activity. Immediate effect was determined based on 7-day windows before and after the visit, and sustained effects were evaluated for weeks 1-8 after module completion. RESULTS Out of the 811 participants, 498 (61.4%) visited the Goal module and 406 (50.1%) visited the Barriers module. The Barriers module had an immediate effect on very active and total active minutes (before-median for very active minutes: 24.2min/day, interquartile range IQR 10.4–43.0min vs. after: 24.9min, IQR 10.0–46.3min; P=.047; before-median for total active minutes: 45.1min/day, interquartile range IQR 22.9–74.9min vs. after: 46.9min, IQR 22.4–78.4min; P=.029). The differences were larger when only completed Barriers modules were considered. Barriers module completion was also associated with sustained effects in fairly active and total active minutes for most of the eight weeks following module completion, and for three weeks in very active minutes. CONCLUSIONS The Barriers module had small significant immediate and sustained effects on active minutes measured by a wrist-worn activity tracker. Future interventions should pay attention to assessing barriers and planning coping mechanisms to overcome them. CLINICALTRIAL ISRTCN registry ISRCTN88405328; https://www.isrctn.com/ISRCTN88405328.