BATframe: An Unsupervised Approach for Domain-Sensitive Affect Detection

Accurate remote sensing of mountainous forest cover change is important for myriad social and ecological reasons, but is challenged by topographic and illumination conditions that can affect detection of forests. Several topographic illumination correction (TIC) approaches have been developed to mitigate these effects, but existing research has focused mostly on whether TIC improves forest cover classification accuracy and has usually found only marginal gains. However, the beneficial effects of TIC may go well beyond accuracy since TIC promises to improve detection of low illuminated forest cover and thereby normalize measurements of the amount, geographic distribution, and rate of forest cover change regardless of illumination. To assess the effects of TIC on the extent and geographic distribution of forest cover change, in addition to classification accuracy, we mapped forest cover across mountainous Nepal using a 25-year (1992–2016) gap-filled Landsat time series in two ways—with and without TIC (i.e., nonTIC)—and classified annual forest cover using a Random Forest classifier. We found that TIC modestly increased classifier accuracy and produced more conservative estimates of net forest cover change across Nepal (−5.2% from 1992–2016) TIC. TIC also resulted in a more even distribution of forest cover gain across Nepal with 3–5% more net gain and 4–6% more regenerated forest in the least illuminated regions. These results show that TIC helped to normalize forest cover change across varying illumination conditions with particular benefits for detecting mountainous forest cover gain. We encourage the use of TIC for satellite remote sensing detection of long-term mountainous forest cover change.

Download Full-text

An Unsupervised Approach to Partial Discharge Monitoring in Rotating Machines: Detection to Diagnosis with Reduced Need of Expert Support

IEEE Transactions on Energy Conversion ◽

10.1109/tec.2021.3050324 ◽

2021 ◽

pp. 1-1

Author(s):

P. Seri ◽

R. Ghosh ◽

G. C. Montanari

Keyword(s):

Partial Discharge ◽

Rotating Machines ◽

Unsupervised Approach ◽

Discharge Monitoring ◽

Expert Support

Download Full-text

CSO Classifier 3.0: a scalable unsupervised method for classifying documents in terms of research topics

International Journal on Digital Libraries ◽

10.1007/s00799-021-00305-y ◽

2021 ◽

Author(s):

Angelo Salatino ◽

Francesco Osborne ◽

Enrico Motta

Keyword(s):

Computer Science ◽

Digital Libraries ◽

Gold Standard ◽

Research Trends ◽

Alternative Methods ◽

Relevant Research ◽

Research Topics ◽

Research Papers ◽

Research Areas ◽

Unsupervised Approach

AbstractClassifying scientific articles, patents, and other documents according to the relevant research topics is an important task, which enables a variety of functionalities, such as categorising documents in digital libraries, monitoring and predicting research trends, and recommending papers relevant to one or more topics. In this paper, we present the latest version of the CSO Classifier (v3.0), an unsupervised approach for automatically classifying research papers according to the Computer Science Ontology (CSO), a comprehensive taxonomy of research areas in the field of Computer Science. The CSO Classifier takes as input the textual components of a research paper (usually title, abstract, and keywords) and returns a set of research topics drawn from the ontology. This new version includes a new component for discarding outlier topics and offers improved scalability. We evaluated the CSO Classifier on a gold standard of manually annotated articles, demonstrating a significant improvement over alternative methods. We also present an overview of applications adopting the CSO Classifier and describe how it can be adapted to other fields.

Download Full-text

A deep learning unsupervised approach for fault diagnosis of household appliances

IFAC-PapersOnLine ◽

10.1016/j.ifacol.2020.12.2856 ◽

2020 ◽

Vol 53 (2) ◽

pp. 10749-10754

Author(s):

Francesco Cordoni ◽

Gianluca Bacchiega ◽

Giulio Bondani ◽

Robert Radu ◽

Riccardo Muradore

Keyword(s):

Deep Learning ◽

Fault Diagnosis ◽

Household Appliances ◽

Unsupervised Approach

Download Full-text

Protein Binding Partners of Dysregulated miRNAs in Parkinson’s Disease Serum

Cells ◽

10.3390/cells10040791 ◽

2021 ◽

Vol 10 (4) ◽

pp. 791

Author(s):

Wolfgang P. Ruf ◽

Axel Freischmidt ◽

Veselin Grozdanov ◽

Valerie Roth ◽

Sarah J. Brockmann ◽

...

Keyword(s):

Parkinson’S Disease ◽

Parkinson's Disease ◽

Protein Binding ◽

Neurodegenerative Diseases ◽

Extracellular Vesicles ◽

Abundance Pattern ◽

Binding Partners ◽

Direct Binding ◽

Unsupervised Approach ◽

High Degree

Accumulating evidence suggests that microRNAs (miRNAs) are a contributing factor to neurodegenerative diseases. Although altered miRNA profiles in serum or plasma have been reported for several neurodegenerative diseases, little is known about the interaction between dysregulated miRNAs and their protein binding partners. We found significant alterations of the miRNA abundance pattern in serum and in isolated serum-derived extracellular vesicles of Parkinson’s disease (PD) patients. The differential expression of miRNA in PD patients was more robust in serum than in isolated extracellular vesicles and could separate PD patients from healthy controls in an unsupervised approach to a high degree. We identified a novel protein interaction partner for the strongly dysregulated hsa-mir-4745-5p. Our study provides further evidence for the involvement of miRNAs and HNF4a in PD. The demonstration that miRNA-protein binding might mediate the pathologic effects of HNF4a both by direct binding to it and by binding to proteins regulated by it suggests a complex role for miRNAs in pathology beyond the dysregulation of transcription.

Download Full-text

An Unsupervised Approach for 3D Medical Image Registration

Proceedings of the 2019 7th International Conference on Information Technology: IoT and Smart City ◽

10.1145/3377170.3377273 ◽

2019 ◽

Author(s):

Yingjun Ma ◽

Jinshuo Zhang ◽

Dongmei Niu ◽

Muhammad Umair Hassan ◽

Xiuyang Zhao

Keyword(s):

Image Registration ◽

Medical Image ◽

Medical Image Registration ◽

Unsupervised Approach

Download Full-text

DNA Methylation-based Signatures Classify Sporadic Pituitary Tumors According to Clinicopathological Features

Neuro-Oncology ◽

10.1093/neuonc/noab044 ◽

2021 ◽

Author(s):

Maritza S Mosella ◽

Thais S Sabedot ◽

Tiago C Silva ◽

Tathiane M Malta ◽

Felipe D Segato ◽

...

Keyword(s):

Functional Status ◽

Validation Cohort ◽

Pituitary Tumors ◽

Clinicopathological Features ◽

Cell Lineages ◽

Adenohypophyseal Cell ◽

Genome Wide ◽

Unsupervised Approach ◽

Recent Classification ◽

Methylation Patterns

Abstract Background Distinct genome-wide methylation patterns cluster pituitary neuroendocrine tumors (PitNETs) into molecular groups associated with specific clinicopathological features. Here we aim to identify, characterize and validate methylation signatures that objectively classify PitNET into clinicopathological groups. Methods Combining in-house and publicly available data, we conducted an analysis of the methylome profile of a comprehensive cohort of 177 tumors (Panpit cohort) and 20 nontumor specimens from the pituitary gland. We also retrieved methylome data from an independent PitNET cohort (N=86) to validate our findings. Results We identified three methylation clusters associated with adenohypophyseal cell lineages and functional status using an unsupervised approach. Differentially methylated CpG probes (DMP) significantly distinguished the Panpit clusters and accurately assigned the samples of the validation cohort to their corresponding lineage and functional subtypes memberships. The DMPs were annotated in regulatory regions enriched of enhancer elements, associated with pathways and genes involved in pituitary cell identity, function, tumorigenesis, invasiveness. Some DMPs correlated with genes with prognostic and therapeutic values in other intra or extracranial tumors. Conclusions We identified and validated methylation signatures, mainly annotated in enhancer regions that distinguished PitNETs by distinct adenohypophyseal cell lineages and functional status. These signatures provide the groundwork to develop an unbiased approach to classifying PitNETs according to the most recent classification recommended by the 2017 WHO and to explore their biological and clinical relevance in these tumors.

Download Full-text