Using transfer learning for diabetic retinopathy stage classification

2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Enas M.F. El Houby
Keyword(s):  
Diabetic Retinopathy ◽  
Transfer Learning ◽  
Grade Level ◽  
Design Methodology ◽  
Content Type ◽  
Stage Classification ◽  
Set Up ◽  
Retinal Fundus ◽  
Automated Methods

PurposeDiabetic retinopathy (DR) is one of the dangerous complications of diabetes. Its grade level must be tracked to manage its progress and to start the appropriate decision for treatment in time. Effective automated methods for the detection of DR and the classification of its severity stage are necessary to reduce the burden on ophthalmologists and diagnostic contradictions among manual readers.Design/methodology/approachIn this research, convolutional neural network (CNN) was used based on colored retinal fundus images for the detection of DR and classification of its stages. CNN can recognize sophisticated features on the retina and provides an automatic diagnosis. The pre-trained VGG-16 CNN model was applied using a transfer learning (TL) approach to utilize the already learned parameters in the detection.FindingsBy conducting different experiments set up with different severity groupings, the achieved results are promising. The best-achieved accuracies for 2-class, 3-class, 4-class and 5-class classifications are 86.5, 80.5, 63.5 and 73.7, respectively.Originality/valueIn this research, VGG-16 was used to detect and classify DR stages using the TL approach. Different combinations of classes were used in the classification of DR severity stages to illustrate the ability of the model to differentiate between the classes and verify the effect of these changes on the performance of the model.

scholarly journals Using Transfer Learning for Diabetic Retinopathy Stages Classification

2021 ◽  
Author(s):  
Enas M.F. El Houby
Keyword(s):  
Diabetic Retinopathy ◽  
Transfer Learning ◽  
Grade Level ◽  
The Body ◽  
Patients At Risk ◽  
Retinal Fundus ◽  
Automated Methods ◽  
Disease Grade ◽  
Visual Geometry

Abstract Purpose – Diabetes is a chronic disease, that leads to damage of many systems of the body. One of the dangerous complications of diabetes is diabetic retinopathy. Frequent inspection for diabetic retinopathy is essential to recognize patients at risk of visual impairment. The disease grade level must be tracked to manage its progress and to start the appropriate decision for treatment in time. Effective automated methods for the detection of diabetic retinopathy and the classification of its severity stage are necessary. It also helps to decrease the burden on ophthalmologists and reduce diagnostic contradictions among manual readers. Methods– In this research, a convolutional neural network (CNN) is used based on color retinal fundus images for the detection of diabetic retinopathy (DR) and classification of its stages. CNN can recognize sophisticated features on the retina and so provide an automatic diagnosis. The pre-trained CNN model Visual Geometry Group (VGG) is applied on DR data using a transfer learning approach to utilize the already learnt parameters based on 1,000,000 images of ImageNet with 1000 classes.Results – By conducting different experiments with different classes setting the built models achieved promising results. The best achieved accuracies for 2-ary, 3-ary, 4-ary, and 5-ary classification are 85.99, 80.5, 61.28, and 71, respectively.

scholarly journals A surrogate-based generic classifier for Chinese TV series reviews

2017 ◽  
Vol 45 (2) ◽  
pp. 66-74
Author(s):  
Yufeng Ma ◽  
Long Xia ◽  
Wenqi Shen ◽  
Mi Zhou ◽  
Weiguo Fan
Keyword(s):  
Design Methodology ◽  
Experimental Result ◽  
Digital Information ◽  
Data Set ◽  
Content Type ◽  
Category Labels ◽  
Information Supply Chain ◽  
Feature Selection Techniques ◽  
Practical Implications

Purpose The purpose of this paper is automatic classification of TV series reviews based on generic categories. Design/methodology/approach What the authors mainly applied is using surrogate instead of specific roles or actors’ name in reviews to make reviews more generic. Besides, feature selection techniques and different kinds of classifiers are incorporated. Findings With roles’ and actors’ names replaced by generic tags, the experimental result showed that it can generalize well to agnostic TV series as compared with reviews keeping the original names. Research limitations/implications The model presented in this paper must be built on top of an already existed knowledge base like Baidu Encyclopedia. Such database takes lots of work. Practical implications Like in digital information supply chain, if reviews are part of the information to be transported or exchanged, then the model presented in this paper can help automatically identify individual review according to different requirements and help the information sharing. Originality/value One originality is that the authors proposed the surrogate-based approach to make reviews more generic. Besides, they also built a review data set of hot Chinese TV series, which includes eight generic category labels for each review.

Viewpoint – could the belt and road initiative be derailed? An analysis of impediments to completion

2020 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Peter Enderwick
Keyword(s):  
Design Methodology ◽  
Dynamic Environment ◽  
Belt And Road Initiative ◽  
Relative Significance ◽  
Content Type ◽  
Belt And Road ◽  
Trade War ◽  
The Belt And Road

Purpose The continuation of China’s belt and road initiative (BRI) is assumed in most analyses. Yet, recent events have created significant reputational damage for China and Chinese businesses. With a trade war evolving into a hegemonic struggle, there are a number of potential developments that could derail the BRI. This paper aims to provide a contemporary review of the factors that could negatively impact its continuation, and what China has done to mitigate the risks. Design/methodology/approach A descriptive paper that groups possible disruptive factors into three groups: internal weaknesses of the BRI and its design; those related to China’s implementation of the BRI and external concerns and pressures. Findings China has actively reviewed and refined the BRI to reduce its perceived weaknesses and increase its attractiveness to potential participants, focussing on debt dependency, transparency and governance. However, this has occurred at the same time as growing concerns regarding China’s international assertiveness, the hegemonic challenge and recovery from the COVID-19 pandemic. Research limitations/implications These changes are occurring within an extremely dynamic environment and any analysis at one point in time is subject to considerable limitations. However, the paper brings together a range of disparate perspectives in a structured manner. Originality/value The classification of possible threats to the BRI is original and provides insights into the relative significance of the diverse challenges that China faces. The paper concludes that while China’s operational focus on the mechanics of the BRI process is necessary, it may not be sufficient to ensure its continuing development. The paper identifies the next step which is conceptualisation of these ideas and of the BRI. Some guidance as to how this might be done is provided.

scholarly journals On classification of (n+6)-dimensional nilpotent n-Lie algebras of class 2 with n≥4

2020 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Mehdi Jamshidi ◽  
Farshid Saeedi ◽  
Hamid Darabi
Keyword(s):  
Lie Algebra ◽  
Lie Algebras ◽  
Design Methodology ◽  
Central Element ◽  
Complete Understanding ◽  
Content Type ◽  
Low Dimension

PurposeThe purpose of this paper is to determine the structure of nilpotent (n+6)-dimensional n-Lie algebras of class 2 when n≥4.Design/methodology/approachBy dividing a nilpotent (n+6)-dimensional n-Lie algebra of class 2 by a central element, the authors arrive to a nilpotent (n+5) dimensional n-Lie algebra of class 2. Given that the authors have the structure of nilpotent (n+5)-dimensional n-Lie algebras of class 2, the authors have access to the structure of the desired algebras.FindingsIn this paper, for each n≥4, the authors have found 24 nilpotent (n+6) dimensional n-Lie algebras of class 2. Of these, 15 are non-split algebras and the nine remaining algebras are written as direct additions of n-Lie algebras of low-dimension and abelian n-Lie algebras.Originality/valueThis classification of n-Lie algebras provides a complete understanding of these algebras that are used in algebraic studies.

Wildcards – natural and artificial: the combination of a panel of experts and Fuzzy TOPSIS

foresight ◽  
2017 ◽  
Vol 19 (1) ◽  
pp. 15-30 ◽  
Author(s):  
Mohsen Mohammadi ◽  
Mohammad Rahim Eivazi ◽  
Jafar Sajjadi
Keyword(s):  
Design Methodology ◽  
Fuzzy Topsis ◽  
New Method ◽  
Weak Signals ◽  
Content Type ◽  
Participatory Foresight ◽  
The Future ◽  
Similar Character ◽  
Topsis Model

Purpose The purpose of this paper is threefold: to classify wildcards into three particular types sharing similar characteristics; use the Fuzzy TOPSIS as a new method in foresight to turn qualitative ideas into quantitative ones; and apply a combination of Fuzzy TOPSIS and a panel of experts to prioritize weak signals. Design/methodology/approach In this paper, the authors classify wildcards into three particular types which share similar character: natural wildcards, artificial wildcards (Degree 1) and artificial wildcards (Degree 2). Wildcards point to unexpected and surprising events including important results that can form watershed in the development of a specific trend. In addition, the authors present a Fuzzy TOPSIS model which can be used in various cases to prioritize a number of weak signals and put them in order, so that the most important ones are likely to yield the wildcard in the future Findings The authors presented a classification of wildcards with the same characteristics being natural wildcards, artificial wildcards (Degree 1) and artificial wildcards (Degree 2). The authors also prioritized the weak signals to deal with the most important ones and take appropriate action in advance so as to minimize possible damages and maximize the benefits of potential wildcards in an uncertain environment. Originality/value In this paper, the authors report on the prioritizing of weak signals by applying Fuzzy TOPSIS and classify wildcards. This is significant because, by identifying the most important weak signals, appropriate actions can be taken in the future if necessary. The paper should be of interest to readers in the area of participatory foresight.

Review of securities class action filings in 2015

2016 ◽  
Vol 17 (3) ◽  
pp. 1-27
Author(s):  
Alexander Aganin
Keyword(s):  
Design Methodology ◽  
Stock Exchange ◽  
Class Action ◽  
Current View ◽  
Content Type ◽  
Industry Sector ◽  
Class Actions ◽  
Trends Over Time ◽  
Loss Index

Purpose To provide an analysis of securities class action filings in 2015 along with related trends over time and a comprehensive current view of the securities class action landscape. Design/methodology/approach Details 2015 securities class actions and related trends in terms of the number and size of filings; market capitalization losses; the litigation exposure of IPOs; the classification of complaints; litigation likelihood for US exchange-listed companies; resolutions (settlements, dismissals or trial verdict outcomes); timing of dismissals and settlements; filing lags; filings against foreign issuers; number of mega filings; recent rulings related to class certification; and concentration of class action activity by industry sector, stock exchange and court circuit. Findings The number of filings in 2015 was the largest since 2008. The Disclosure Dollar Loss Index® (DDL Index®), the Maximum Dollar Loss Index® (MDL Index®) and the number of mega filings rose sharply in 2015 after declines in 2014. The Consumer Non-Cyclical sector had the most filings in 2015 while filings against companies in the Financial sector were below historical averages. Dismissal rates appear to be trending down. The median filing lag has never been shorter than in 2015. Filings against foreign issuers remain at high levels. Filings against S&P 500 companies remained below the historical average. Originality/value Detailed analysis by legal and industry experts.

Excellence and gender

2015 ◽  
Vol 34 (1) ◽  
pp. 37-54 ◽  
Author(s):  
Farinaz Fassa
Keyword(s):  
Gender Equality ◽  
Common Good ◽  
Design Methodology ◽  
Local Level ◽  
National Level ◽  
Academic Careers ◽  
Content Type ◽  
And Gender ◽  
Set Up ◽  
Feminist Movements

Purpose – Discussing the Swiss case, the purpose of this paper is to examine how gender equality policies deal with the present requirements for scholars to be considered “excellent”. It aims to pinpoint the lines of tension or coherence between excellence, meritocracy and gender politics. Design/methodology/approach – In order to specify the norms of academic careers and their different renditions, the author draws on two studies (at local and national levels) to illustrate where the changes and resistances are taking place. Findings – The translations of a number of demands of feminist movements into the policies set up to favour equality between the sexes may combine to challenge the norms of academia as a gendered realm. Nevertheless, without strong pressure from feminists at local level and the conduct of research pursuing the enterprise of deconstructing norms, top-down policies may prove less “corrective” than affirmative action. This pressure is not only useful to build gender equality in science but also to broaden the spectrum of knowledge that can become a common good. Research limitations/implications – As neither the names nor the positions of the experts who select the candidates at national level are made public, we had to opt for other, less satisfactory means. Originality/value – The originality of the paper lies in the link made between the enhancements brought by Equalities policies and the changes they bring. It attempts to bring to light the extent to which gender equality policies conform to the neo-managerial order or challenge its norms to build a world that is more just.

scholarly journals An economic theory of Islamic finance

2017 ◽  
Vol 9 (2) ◽  
pp. 117-132 ◽  
Author(s):  
Mabid Ali Al-Jarhi
Keyword(s):  
Design Methodology ◽  
Islamic Finance ◽  
Economic Systems ◽  
Monetary Reform ◽  
Content Type ◽  
Reform Agenda ◽  
Banking And Finance ◽  
Economic Rationale ◽  
Islamic Economic ◽  
Set Up

Purpose This paper aims to provide an economic rationale for Islamic finance. Design/methodology/approach Its methodology is simple. It starts with listing the contributions to economic analysis relevant to the required rationale in the theories of banking, finance, price, money and macroeconomics, to identify the main rationale for Islamic finance. A concise description of the author’s model for an Islamic economic system, within which Islamic finance can be operational, is provided. Findings The paper finds distinct advantages of Islamic finance, when properly applied within the author’s model. Islamic finance can therefore be a candidate as a reform agenda for conventional finance. It opens the door for significant monetary reform in currently prevalent economic systems. Research limitations/implications The first limitation of the paper is that the distinct benefits of Islamic finance are all of macroeconomic types which are external to Islamic banking and finance institutions. They are therefore not expected to motivate such institutions to apply Islamic finance to the letter, without regulators interference to ensure strict application. The second limitation is the necessity to set up enabling institutional and regulatory arrangements for Islamic finance. Originality/value The results are unique as they challenge the received doctrine and provide non-religious rationale for Islamic finance.

Sign in / Sign up

Export Citation Format

Share Document