Filters








68 Hits in 2.6 sec

Unsupervised Progressive Learning and the STAM Architecture [article]

James Smith, Cameron Taylor, Seth Baer, Constantine Dovrolis
2021 arXiv   pre-print
We first pose the Unsupervised Progressive Learning (UPL) problem: an online representation learning problem in which the learner observes a non-stationary and unlabeled data stream, learning a growing  ...  While there are no existing learning scenarios that are directly comparable to UPL, we compare the STAM architecture with two recent continual learning models, Memory Aware Synapses (MAS) and Gradient  ...  Acknowledgements This work is supported by the Lifelong Learning Machines (L2M) program of DARPA/MTO: Cooperative Agreement HR0011-18-2-0019.  ... 
arXiv:1904.02021v6 fatcat:2dqxyms5kffgpoxbxryy4umskm

Unsupervised Progressive Learning and the STAM Architecture

James Smith, Cameron Taylor, Seth Baer, Constantine Dovrolis
2021 Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence   unpublished
We first pose the Unsupervised Progressive Learning (UPL) problem: an online representation learning problem in which the learner observes a non-stationary and unlabeled data stream, learning a growing  ...  While there are no existing learning scenarios that are directly comparable to UPL, we compare the STAM architecture with two recent continual learning models, Memory Aware Synapses (MAS) and Gradient  ...  Figure 7 : 7 Ablation study: A STAM architecture without LTM (left), a STAM architecture in which the LTM centroids are adjusted with the same learning rate α as in STM (center), and a STAM architecture  ... 
doi:10.24963/ijcai.2021/410 fatcat:cjj3k75eoffhhmtilskofi7k4y

CDN-MEDAL: Two-stage Density and Difference Approximation Framework for Motion Analysis [article]

Synh Viet-Uyen Ha, Cuong Tien Nguyen, Hung Ngoc Phan, Nhat Minh Chung, Phuong Hoai Ha
2021 arXiv   pre-print
The first architecture, CDN-GM, is grounded on an unsupervised GMM statistical learning strategy to describe observed scenes' salient features.  ...  However, the techniques have only provided limited descriptions of scenes' properties while requiring heavy computations, as their single-valued mapping functions are learned to approximate the temporal  ...  , DeepBS, STAM, FgSegNet, and Cascade CNN.  ... 
arXiv:2106.03776v4 fatcat:hoiazfi4zfgf3hxv7ybiwinp5e

Unsupervised Continual Learning via Self-Adaptive Deep Clustering Approach [article]

Mahardhika Pratama, Andri Ashfahani, Edwin Lughofer
2021 arXiv   pre-print
Unsupervised continual learning remains a relatively uncharted territory in the existing literature because the vast majority of existing works call for unlimited access of ground truth incurring expensive  ...  Another issue lies in the problem of task boundaries and task IDs which must be known for model's updates or model's predictions hindering feasibility for real-time deployment.  ...  Learn-to-grow is proposed in [Li et al., 2019] where it utilizes the neural architecture search to find the best network structure of a given task.  ... 
arXiv:2106.14563v1 fatcat:5h3rimq76bcaxgq6gafml6ih3u

TensorMoG: A Tensor-Driven Gaussian Mixture Model with Dynamic Scene Adaptation for Background Modelling

Synh Viet-Uyen Ha, Nhat Minh Chung, Hung Ngoc Phan, Cuong Tien Nguyen
2020 Sensors  
Therefore, this paper proposes an unsupervised, parallelized, and tensor-based approach that algorithmically works with entropy estimations.  ...  takes into account those effects and distorts the results.  ...  We would also like to express my sincere gratitude to anonymous insightful reviewers whose comments provided pearls of wisdom during the course of this research to help us improve and clarify this manuscript  ... 
doi:10.3390/s20236973 pmid:33291320 pmcid:PMC7729891 fatcat:wjuppaq5drgwnee6bqy4evtpiu

stam--a Bioconductor compliant R package for structured analysis of microarray data

Claudio Lottaz, Rainer Spang
2005 BMC Bioinformatics  
This is done using both gene expression data and functional gene annotations. We provide stam, a Bioconductor compliant software package for the statistical programming environment R.  ...  Furthermore, stam finds novel sub-group stratifications of patients according to the absence or presence of molecular symptoms.  ...  Hagemeier and Karl Seeger from the Charité Medical Center for fruitful discussions.  ... 
doi:10.1186/1471-2105-6-211 pmid:16122395 pmcid:PMC1208856 fatcat:7uug3xgmqzazbmhfk7u6ipyaq4

DAWN: Dual Augmented Memory Network for Unsupervised Video Object Tracking [article]

Zhenmei Shi, Haoyang Fang, Yu-Wing Tai, Chi-Keung Tang
2019 arXiv   pre-print
Psychological studies have found that human visual tracking system involves learning, memory, and planning.  ...  Despite recent successes, not many works have focused on memory and planning in deep learning based tracking.  ...  Acknowledgement This research is supported by Tencent and the Research Grant Council of Hong Kong SAR under grant no. 16201818.  ... 
arXiv:1908.00777v2 fatcat:xqqcajz4i5hwrecv3o2zvx4cke

Representation learning for improved interpretability and classification accuracy of clinical factors from EEG [article]

Garrett Honke, Irina Higgins, Nina Thigpen, Vladimir Miskovic, Katie Link, Sunny Duan, Pramod Gupta, Julia Klawohn, Greg Hajcak
2020 arXiv   pre-print
Here we adapt an unsupervised pipeline from the recent deep representation learning literature to address these problems by 1) learning a disentangled representation using β-VAE to denoise the signal,  ...  Finally, single factors of the learned disentangled representations often correspond to meaningful markers of clinical factors, as automatically detected by SCAN, allowing for human interpretability and  ...  Recently great progress has been made in the field of deep unsupervised representation learning (Roy et al., 2019; Devlin et al., 2018; Brown et al., 2020; Chen et al., 2020b; Grill et al., 2020; Chen  ... 
arXiv:2010.15274v3 fatcat:vnhzn6j7ybgnlj2ccvh5qqcuzm

A system design to support outside activities of older adults using smart urban objects

Julian Fietkau, Laura Stojko
2020 European Conference on Computer Supported Cooperative Work  
an activity and receives navigational assistance to increase their motivation and feeling of safety while undertaking the chosen activity.  ...  Finally, we discuss our approach regarding challenges such as user autonomy, privacy and real-world deployments, which need to be considered in future implementation and evaluation phases of the system  ...  Acknowledgments This work has been supported by the Federal Ministry of Education and Research, Germany, under grant 16SV7443. We thank all project partners for their commitment.  ... 
doi:10.18420/ecscw2020_ep07 dblp:conf/ecscw/FietkauS20 fatcat:4mzkg26jwbf73k2qeemeq2vk2a

State-space model with deep learning for functional dynamics estimation in resting-state fMRI

Heung-Il Suk, Chong-Yaw Wee, Seong-Whan Lee, Dinggang Shen
2016 NeuroImage  
Specifically, we propose a novel methodological architecture that combines deep learning and state-space modelling, and apply it to rs-fMRI based Mild Cognitive Impairment (MCI) diagnosis.  ...  We also analyzed the functional networks learned by DAE, estimated the functional connectivities by decoding hidden states in HMM, and investigated the estimated functional connectivities by means of a  ...  B0101-15-0307, Basic Software Research in Human-level Lifelong Machine Learning (Machine Learning Center)).  ... 
doi:10.1016/j.neuroimage.2016.01.005 pmid:26774612 pmcid:PMC5437848 fatcat:b5fhyiv7lzb5zegq2jptwhj7ve

Stacked Generative Machine Learning Models for Fast Approximations of Steady-State Navier-Stokes Equations [article]

Shen Wang, Mehdi Nikfar, Joshua C. Agar, Yaling Liu
2021 arXiv   pre-print
To improve computational efficiency, machine learning techniques have been used to create accelerated data-driven approximations for CFD.  ...  We achieve state-of-the-art results without any labeled simulation data, but using a custom data-driven and physics-informed loss function by using and small-scale solutions to prime the model to solve  ...  Small Data Challenges in Big Data Era: A Survey of Recent Progress on Unsupervised and Semi-Supervised Methods. en. IEEE Trans. Pattern Anal. Mach. Intell. PP (Oct. 2020). 24.  ... 
arXiv:2112.06419v1 fatcat:hd4caoonive5dokqz7x2cqqq7u

Bridging Scales in Alzheimer's Disease: Biological Framework for Brain Simulation With The Virtual Brain

Leon Stefanovski, Jil Mona Meier, Roopa Kalsank Pai, Paul Triebkorn, Tristram Lett, Leon Martin, Konstantin Bülau, Martin Hofmann-Apitius, Ana Solodkin, Anthony Randal McIntosh, Petra Ritter
2021 Frontiers in Neuroinformatics  
Alzheimer's Disease (AD) is the most common cause of dementia and represents the most prevalent neurodegenerative disease.  ...  Despite the acceleration of knowledge and data accumulation in neuroscience over the last years, the highly prevalent neurodegenerative disease of AD remains a growing problem.  ...  This disconnection correlated with the cognitive and behavioral decline (Stam, 2014) and white matter pathology in certain areas could be used as a biomarker for disease progression (Solodkin et al.  ... 
doi:10.3389/fninf.2021.630172 pmid:33867964 pmcid:PMC8047422 fatcat:khujbqtyxfhwvk66l4azu7hg54

Automatic Analysis of EEGs Using Big Data and Hybrid Deep Learning Architectures

Meysam Golmohammadi, Amir Hossein Harati Nejad Torbati, Silvia Lopez de Diego, Iyad Obeid, Joseph Picone
2019 Frontiers in Human Neuroscience  
This hybrid architecture integrates hidden Markov models (HMMs) for sequential decoding of EEG events with deep learning-based post-processing that incorporates temporal and spatial context.  ...  In this study, a high-performance automated EEG analysis system based on principles of machine learning and big data is proposed.  ...  Steven Tobochnik and David Jungries of the Temple University School of Medicine for their assistance in developing the classification paradigm used in this study and for preparing the manually annotated  ... 
doi:10.3389/fnhum.2019.00076 pmid:30914936 pmcid:PMC6423064 fatcat:tptts5ky6jftbcj7ankxqaenoe

Informative Neural Codes to Separate Object Categories [article]

Mozhgan Shahmohammadi, Ehsan Vahab, Hamid Karimi-Rouzbahani
2020 bioRxiv   pre-print
Among the evaluated features, event-related potential (ERP) components of N1 and P2a were among the most in-formative features with the highest information in the Theta frequency bands.  ...  This has already inspired AI-based object recognition algorithms, such as convolutional neural networks, which are among the most successful object recognition platforms today and can approach human performance  ...  They are categorized into one of the two categories of "machine learning" approaches and "deep learning" approaches.  ... 
doi:10.1101/2020.12.04.409789 fatcat:dq47iobvdjbhxa5uvkul5udb2y

Automatic Analysis of EEGs Using Big Data and Hybrid Deep Learning Architectures [article]

Meysam Golmohammadi, Amir Hossein Harati Nejad Torbati, Silvia Lopez de Diego, Iyad Obeid, Joseph Picone
2017 arXiv   pre-print
These algorithms were trained and evaluated using the TUH EEG Corpus, which is the world's largest publicly available database of clinical EEG data.  ...  We propose a highperformance classification system based on principles of big data and machine learning.  ...  ACKNOWLEDGEMENTS The primary funder of this research was the QED Proof of Concept program of the University City Science Center (Grant No. S1313).  ... 
arXiv:1712.09771v1 fatcat:63swak52kzcy5db64hzp47shhm
« Previous Showing results 1 — 15 out of 68 results