Abstract Objective: To determine how well machine learning algorithms can classify mild cognitive impairment (MCI) subtypes and Alzheimer’s disease (AD) using features obtained from the digital Clock Drawing Test (dCDT). Methods: dCDT protocols were administered to 163 patients diagnosed with AD( n = 59), amnestic MCI (aMCI; 26), combined mixed/dysexecutive (mixed/dys MCI; 43), without (non-MCI; 35) standard clock drawing command copy procedures, that is, draw face of clock, put in all...

E-learning can be loosely defined as a wide set of applications and processes, which uses available electronic media (and tools) to deliver vocational education training. With its increasing recognition an ubiquitous mode instruction interaction in the academic well corporate world, need for scaleable realistic model is becoming important. In this paper, we introduce SELF; semantic grid-based e-learning framework. SELF aims identify key-enablers practical environment minimize technological...

Process anomaly is used for detecting cyber-physical attacks on critical infrastructure such as plants water treatment and electric power generation. Identification of process possible using rules that govern the physical chemical behavior within a plant. These rules, often referred to invariants, can be derived either directly from plant design or data generated in an operational. However, operational legacy plants, one might consider data-centric approach derivation invariants. The study...

One of the more challenging real-world problems in computational intelligence is to learn from non-stationary streaming data, also known as concept drift. Perhaps even a version this scenario when - following small set initial labeled data stream consists unlabeled only. Such typically referred learning initially nonstationary environment, or simply extreme verification latency (EVL). In our prior work, we described framework, called COMPOSE (COMPacted Object Sample Extraction) that works...

Adversarial learning is used to test the robustness of machine algorithms under attack and create attacks that deceive anomaly detection methods in Industrial Control System (ICS). Given security assessment an ICS demands exhaustive set possible patterns studied, this work, we propose association rule mining-based generation technique. The technique has been implemented using data from a Secure Water Treatment plant. proposed was able generate more than 110,000 constituting vast majority new...

With the advancement of mobile technologies, pedagogical innovations have been emerging since last two decades. The potential use devices has attracted many educators and researchers to conduct inquiry based learning (IBL) activities in a classroom or fieldwork. In recent years, augmented reality (AR) technology become widespread for conducting using devices. literature related science education, there are limited systems found which AR is used IBL activities. We therefore designed system...

Intelligent transportation systems enables the analysis of large multidimensional street traffic data to detect pattern and anomaly, which otherwise is a difficult task. Advancement in computer vision makes great contribution progress video based surveillance system. But still there are some challenges need be solved like objects occlusion, behavior objects. This paper developed novel framework explores road analyze different patterns anomaly detection. implemented on dataset collected from...

Artificial neural networks are well-known to be susceptible catastrophic forgetting when continually learning from sequences of tasks. Various continual (or "incremental") approaches have been proposed avoid forgetting, but they typically adversary agnostic, i.e., do not consider the possibility a malicious attack. In this effort, we explore vulnerability Elastic Weight Consolidation (EWC), popular algorithm for avoiding forgetting. We show that an intelligent can take advantage EWC's...

E-learning can be loosely defined as a wide set of applications and processes, which uses available electronic media (and tools) to deliver vocational education training. With its increasing recognition an ubiquitous mode instruction interaction in the academic well corporate world, need for scaleable realistic model is becoming important. In this paper we introduce SELF; Semantic grid-based E-Learning Framework. SELF aims identify key-enablers practical environment minimize technological...

Nonstationary streaming data are characterized by changes in the underlying distribution between subsequent time steps. Learning such environments becomes even more challenging when labeled available only at initial step, and algorithm is provided unlabeled thereafter, a scenario referred to as extreme verification latency. Our previously introduced COMPOSE framework works very well settings. semi-supervised approach that iteratively labels strategically chosen instances of next step using...

Domain adaptation techniques such as importance weighting modify the training data to better represent a different test distribution, process that may be particularly vulnerable malicious attack in an adversarial machine learning scenario. In this work, we explore level of vulnerability poisoning attacks. Importance weighting, like other domain approaches, assumes distributions and are but related. An intelligent adversary, having full or partial access data, can take advantage expected...

Adversarial machine learning has recently risen to prominence due increased concerns over the vulnerability of algorithms malicious attacks. While impact poisoning attacks on some popular algorithms, such as deep neural networks, been well researched, other approaches not yet properly established. In this effort, we explore unconstrained least squares importance fitting (uLSIF), an algorithm used for computing ratio covariate shift domain adaptation problems. The uLSIF is accurate and...

Cyber-Physical Systems (CPS) have gained popularity due to the increased requirements on their uninterrupted connectivity and process automation. Due over network including intranet internet, dependence sensitive data, heterogeneous nature, large-scale deployment, they are highly vulnerable cyber-attacks. Cyber-attacks performed by creating anomalies in normal operation of systems with a goal either disrupt or destroy system completely. The study proposed here focuses detecting those which...

A mobile edge computing (MEC) server-based scheme is proposed and experiment in this paper order to support the network edge-based video adaptation on MPEG-DASH. Furthermore, emphasize of understand challenges UHD 4K conference 5G that business organizations individual are facing. model based MEC server Caching replacement strategy designed implemented. To facilitate fast streaming conferencing through DASH it needed cache popular content segments enhance performance via caching strategy....

Continual (or "incremental") learning approaches are employed when additional knowledge or tasks need to be learned from subsequent batches streaming data. However these typically adversary agnostic, i.e., they do not consider the possibility of a malicious attack. In our prior work, we explored vulnerabilities Elastic Weight Consolidation (EWC) perceptible misinformation. We now explore other regularization-based as well generative replay-based continual algorithms, and also extend attack...