Application of Modified PageRank Algorithm for Anomaly Detection in Movements of Older Adults

Payandeh, Shahram; Chiu, Eddie

doi:10.1155/2019/8612021

Cited by 6 publications

(3 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Let us assume that for a given window of observa-tion (i.e. time of the day, period of recording, or the number of recording periods over a week or months (Payandeh et al, 2019a;Payandeh et al, 2019b), we can determine the movement distributions (or state propagation) between various areas (or states). These distributions are represented by the directional arrows pointing between various locations (i.e.…”

Section: Full Markov Models (Mm)mentioning

confidence: 99%

An overview of Markov Chains for monitoring indoor movements and ambient motion detection

Payandeh¹

2022

Gerontechnology

Self Cite

View full text Add to dashboard Cite

Background:In recent years and due to the availability of non-wearable sensing technologies, there has been a widespread interest in developing methods for monitoring the movements of older adults and further estimating any anomalies. Objective: Low-cost motion detection has been gaining particular attention due to its widespread availability, ease of deployment, and inherent property in protecting privacy. In addition, and due to the cost-benefits, it has become an attractive alternative for the low-income sector of society. This paper presents an overview of a monitoring framework and a review of the literature using motion detection sensors. Method: Markov Chains (MC) have been explored by many researchers as a suitable framework for monitoring and estimating sequences of events associated with movements and activities which can further be used as a part of an anomaly detection system in a sensor network. This paper presents an overview of this method and related literature. Results: A brief overview of MC and the related literature with some insights and challenges associated with the potential limitations and future extensions. One of the challenges of utilizing MC is the definition of what can be considered the state of movements and activities and what can be used as a measure of such state. In this context, various extensions of MC have been utilized where the state of the system can not be measured directly and are defined in a form of Hidden Markov Models (HMM). Conclusion: Proper deployment of motion detection sensors and associated MC for monitoring requires an in-depth understanding and co-designing of the system with the family members, care providers, and engineers in order to fully take advantage of such technology. Depending on the number and properties of the selected sensors (such as their effective range and the inherent time delay), particular attention needs to be paid to what can constitute a measurable (observable) state and what can be further defined as a hidden state.

show abstract

Section: Full Markov Models (Mm)mentioning

confidence: 99%

An overview of Markov Chains for monitoring indoor movements and ambient motion detection

Payandeh¹

2022

Gerontechnology

Self Cite

View full text Add to dashboard Cite

show abstract

“…Several techniques specifically designed for large-scale ranking problems [34,48] have been used to explore large datasets of images [27] and shapes [16]. PageRank-based approaches in particular have been popular [27,49,56] due to their scalability. [31] proposed an algorithm for object discovery that combines appearance and geometric consistency with PageRank-based link analysis for category discovery.…”

Section: Related Workmentioning

confidence: 99%

Large-Scale Unsupervised Object Discovery

Vo¹,

Sizikova²,

Schmid³

et al. 2021

Preprint

View full text Add to dashboard Cite

Existing approaches to unsupervised object discovery (UOD) do not scale up to large datasets without approximations which compromise their performance. We propose a novel formulation of UOD as a ranking problem, amenable to the arsenal of distributed methods available for eigenvalue problems and link analysis. Extensive experiments with COCO [42] and OpenImages [35] demonstrate that, in the single-object discovery setting where a single prominent object is sought in each image, the proposed LOD (Large-scale Object Discovery) approach is on par with, or better than the state of the art for medium-scale datasets (up to 120K images), and over 37% better than the only other algorithms capable of scaling up to 1.7M images. In the multi-object discovery setting where multiple objects are sought in each image, the proposed LOD is over 14% better in average precision (AP) than all other methods for datasets ranging from 20K to 1.7M images. 1 Figure 1: Sample UOD results obtained by LOD on the OpenImages dataset [35] which contains 1.7M images. Ground-truth boxes are shown in yellow, and predictions are in red. Best viewed in color. 1 Code will be released at https://github.com/huyvvo/LOD Preprint. Under review.

show abstract

“…PageRank in e-commerce system allows users to get the appropriate recommendation of products [9]. To analyze human movement, motor sensors are used to identify the movement path based on the importance ranking of sensors calculated by PageRank [10]. TextRank algorithm was developed to extract keywords by obtaining ranking for a text array based on PageRank algorithm [11].…”

Section: Introductionmentioning

confidence: 99%

Multi-Mode SpMV Accelerator for Transprecision PageRank With Real-World Graphs

Kim

Lee

et al. 2023

IEEE Access

View full text Add to dashboard Cite

With the development of Internet networks, the PageRank algorithm, which was initially developed to recommend important pages in Google's web search systems, is widely used as the basis of various ranking systems in graph processing fields. However, PageRank algorithm requires Sparse Matrix-Vector Multiplication (SpMV) repeatedly which becomes main bottleneck for the calculation. In this study, we present the multi-mode SpMV accelerator for half-to-single transprecision PageRank with real-world graphs. To support transprecision, where the operation performs in half-precision (FP16) initially and changes its precision to single-precision (FP32), the proposed multi-mode SpMV accelerator can perform both dual FP16 mode and single FP32 mode. In dual FP16 mode, the proposed accelerator performs two FP16 SpMV in parallel, and in single FP32 mode, it performs one FP32 SpMV with the same hardware resources. Also, for the reduction of memory footprint, the proposed accelerator supports the Compressed Sparse Row (CSR) format. In addition, dual-issue accumulator and multi-mode transprecision multiplier are presented to support both FP16 and FP32 modes. Validation of the proposed transprecision PageRank algorithm is performed with four real-world graph datasets, resulting in a low 0-4% error rate and 1.3×-1.9× speedup compared to single-precision PageRank computation without transprecision.

show abstract

Application of Modified PageRank Algorithm for Anomaly Detection in Movements of Older Adults

Cited by 6 publications

References 12 publications

An overview of Markov Chains for monitoring indoor movements and ambient motion detection

An overview of Markov Chains for monitoring indoor movements and ambient motion detection

Large-Scale Unsupervised Object Discovery

Multi-Mode SpMV Accelerator for Transprecision PageRank With Real-World Graphs

Contact Info

Product

Resources

About