scholarly journals Deep Learning, Feature Learning, and Clustering Analysis for SEM Image Classification

2020 ◽  
Vol 2 (4) ◽  
pp. 513-528
Author(s):  
Rossella Aversa ◽  
Piero Coronica ◽  
Cristiano De Nobili ◽  
Stefano Cozzini
Keyword(s):  
Unsupervised Learning ◽  
Clustering Analysis ◽  
Feature Learning ◽  
Machine Learning Algorithms ◽  
Learning Approaches ◽  
Clustering Methods ◽  
Data Set ◽  
Sem Image ◽  
Supervised And Unsupervised Learning ◽  
Scanning Electron

In this paper, we report upon our recent work aimed at improving and adapting machine learning algorithms to automatically classify nanoscience images acquired by the Scanning Electron Microscope (SEM). This is done by coupling supervised and unsupervised learning approaches. We first investigate supervised learning on a ten-category data set of images and compare the performance of the different models in terms of training accuracy. Then, we reduce the dimensionality of the features through autoencoders to perform unsupervised learning on a subset of images in a selected range of scales (from 1 μm to 2 μm). Finally, we compare different clustering methods to uncover intrinsic structures in the images.

scholarly journals A novel Approach Using “Supervised and Unsupervised learning” to prevent the Adequacy of Intrusion Detection Systems

2018 ◽  
Vol 7 (3.34) ◽  
pp. 474
Author(s):  
Pradeep Kumar Mallick ◽  
Bibhu Prasad Mohanty ◽  
Sudan Jha ◽  
Kuhoo .
Keyword(s):  
Intrusion Detection ◽  
Unsupervised Learning ◽  
Supervised Learning ◽  
Critical Role ◽  
Intrusion Detection Systems ◽  
Detection Methods ◽  
The Internet ◽  
Learning Approaches ◽  
Detection Systems ◽  
Supervised And Unsupervised Learning

Countering digital dangers, particularly assault detection, is a testing region of research in the field of data affirmation. Intruders utilize polymorphic instruments to disguise the assault payload and dodge the detection methods. Many supervised and unsupervised learning comes closer from the field of machine learning and example acknowledgments have been utilized to expand the adequacy of intrusion detection systems (IDSs). Supervised learning approaches utilize just marked examples to prepare a classifier, however getting adequate named tests is lumbering, and requires the endeavors of area specialists. Notwithstanding, un-marked examples can without much of a stretch be acquired in some genuine issues. Contrasted with super-vised learning approaches, semi-supervised learning (SSL) addresses this issue by considering expansive number of unlabeled examples together with the marked examples to fabricate a superior classifier. In today’s age security is a big issue and every day when we are on the internet we are exposed to a huge number of threats where our personal information can be leaked. The information security and the Intrusion Detection System (IDS) play a critical role in the internet. IDS isan essential tool for detecting different kinds of attacks in a network and maintaining data integrity, confidentiality, and system availability against possible threats. In this paper, we are proposing a modified Elitist approach where the value of fitness is multiplied by the times a variable which is determined on the basis of the value of Kappa (K).  

A machine learning framework for enhancing digital experiences in cultural heritage

2020 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Abdelhak Belhi ◽  
Abdelaziz Bouras ◽  
Abdulaziz Khalid Al-Ali ◽  
Sebti Foufou
Keyword(s):  
Machine Learning ◽  
Cultural Heritage ◽  
Unsupervised Learning ◽  
Research Work ◽  
Data Set ◽  
Content Type ◽  
Data Enrichment ◽  
Supervised And Unsupervised Learning ◽  
Visual Reconstruction ◽  
Cultural Data

PurposeDigital tools have been used to document cultural heritage with high-quality imaging and metadata. However, some of the historical assets are totally or partially unlabeled and some are physically damaged, which decreases their attractiveness and induces loss of value. This paper introduces a new framework that aims at tackling the cultural data enrichment challenge using machine learning.Design/methodology/approachThis framework focuses on the automatic annotation and metadata completion through new deep learning classification and annotation methods. It also addresses issues related to physically damaged heritage objects through a new image reconstruction approach based on supervised and unsupervised learning.FindingsThe authors evaluate approaches on a data set of cultural objects collected from various cultural institutions around the world. For annotation and classification part of this study, the authors proposed and implemented a hierarchical multimodal classifier that improves the quality of annotation and increases the accuracy of the model, thanks to the introduction of multitask multimodal learning. Regarding cultural data visual reconstruction, the proposed clustering-based method, which combines supervised and unsupervised learning is found to yield better quality completion than existing inpainting frameworks.Originality/valueThis research work is original in sense that it proposes new approaches for the cultural data enrichment, and to the authors’ knowledge, none of the existing enrichment approaches focus on providing an integrated framework based on machine learning to solve current challenges in cultural heritage. These challenges, which are identified by the authors are related to metadata annotation and visual reconstruction.

scholarly journals A clustering method of Gas load based on FCM-SMOTE

2021 ◽  
Vol 257 ◽  
pp. 01032
Author(s):  
Dong Hong Huang ◽  
Dan Liu ◽  
Ming Wen ◽  
Xin Li Dong ◽  
Min Wen ◽  
...  
Keyword(s):  
Cluster Analysis ◽  
Clustering Analysis ◽  
Unbalanced Data ◽  
Data Sets ◽  
Current Load ◽  
Clustering Methods ◽  
Clustering Method ◽  
Data Set ◽  
Actual Application ◽  
Data Objects

For the design and planning of gas-fired boiler system, the load of gas-fired boiler is an important basic data. Load clustering analysis, combined with the application of data mining technology and gas boiler system, excavates the hidden load patterns in a large number of disordered and irregular loads, and classifies them, so as to solve many problems in gas boiler system. The current load clustering methods have more or less problems. The invention first carries out data PVA dimension reduction processing on the huge gas data, and then carries out cluster analysis. In the actual application of gas-fired boilers, the data objects we are faced with are usually unbalanced data sets. In order to solve the problem of sample imbalance, we use the FCM-SMOTE algorithm to oversample the clustered data to make the data set into a balanced data set.

Sign in / Sign up

Export Citation Format

Share Document