scholarly journals Identifying Collapsed Buildings Using Post-Earthquake Satellite Imagery and Convolutional Neural Networks: A Case Study of the 2010 Haiti Earthquake

2018 ◽  
Vol 10 (11) ◽  
pp. 1689 ◽  
Author(s):  
Min Ji ◽  
Lanfa Liu ◽  
Manfred Buchroithner
Keyword(s):  
Satellite Imagery ◽  
Human Life ◽  
Haiti Earthquake ◽  
Minority Class ◽  
Imbalance Problem ◽  
Kappa Value ◽  
Under Sampling ◽  
Collapsed Buildings ◽  
Imbalance Learning ◽  
The Cost

Earthquake is one of the most devastating natural disasters that threaten human life. It is vital to retrieve the building damage status for planning rescue and reconstruction after an earthquake. In cases when the number of completely collapsed buildings is far less than intact or less-affected buildings (e.g., the 2010 Haiti earthquake), it is difficult for the classifier to learn the minority class samples, due to the imbalance learning problem. In this study, the convolutional neural network (CNN) was utilized to identify collapsed buildings from post-event satellite imagery with the proposed workflow. Producer accuracy (PA), user accuracy (UA), overall accuracy (OA), and Kappa were used as evaluation metrics. To overcome the imbalance problem, random over-sampling, random under-sampling, and cost-sensitive methods were tested on selected test A and test B regions. The results demonstrated that the building collapsed information can be retrieved by using post-event imagery. SqueezeNet performed well in classifying collapsed and non-collapsed buildings, and achieved an average OA of 78.6% for the two test regions. After balancing steps, the average Kappa value was improved from 41.6% to 44.8% with the cost-sensitive approach. Moreover, the cost-sensitive method showed a better performance on discriminating collapsed buildings, with a PA value of 51.2% for test A and 61.1% for test B. Therefore, a suitable balancing method should be considered when facing imbalance dataset to retrieve the distribution of collapsed buildings.

scholarly journals Discrimination of Earthquake-Induced Building Destruction from Space Using a Pretrained CNN Model

2020 ◽  
Vol 10 (2) ◽  
pp. 602 ◽  
Author(s):  
Min Ji ◽  
Lanfa Liu ◽  
Rongchun Zhang ◽  
Manfred F. Buchroithner
Keyword(s):  
Neural Network ◽  
High Resolution ◽  
Satellite Imagery ◽  
Human Life ◽  
Haiti Earthquake ◽  
Intermediate Layers ◽  
High Resolution Satellite Imagery ◽  
Collapsed Buildings ◽  
Very High ◽  
Visual Geometry

The building is an indispensable part of human life which provides a place for people to live, study, work, and engage in various cultural and social activities. People are exposed to earthquakes, and damaged buildings caused by earthquakes are one of the main threats. It is essential to retrieve the detailed information of affected buildings after earthquakes. Very high-resolution satellite imagery plays a key role in retrieving building damage information since it captures imagery quickly and effectively after the disaster. In this paper, the pretrained Visual Geometry Group (VGG)Net model was applied for identifying collapsed buildings induced by the 2010 Haiti earthquake using pre- and post-event remotely sensed space imagery, and the fine-tuned pretrained VGGNet model was compared with the VGGNet model trained from scratch. The effects of dataset augmentation and freezing different intermediate layers were also explored. The experimental results demonstrated that the fine-tuned VGGNet model outperformed the VGGNet model trained from scratch with increasing overall accuracy (OA) from 83.38% to 85.19% and Kappa from 60.69% to 67.14%. By taking advantage of dataset augmentation, OA and Kappa went up to 88.83% and 75.33% respectively, and the collapsed buildings were better recognized with a larger producer accuracy of 86.31%. The present study showed the potential of using the pretrained Convolutional Neural Network (CNN) model to identify collapsed buildings caused by earthquakes using very high-resolution satellite imagery.

scholarly journals SMOTE: Synthetic Minority Over-sampling Technique

2002 ◽  
Vol 16 ◽  
pp. 321-357 ◽  
Author(s):  
N. V. Chawla ◽  
K. W. Bowyer ◽  
L. O. Hall ◽  
W. P. Kegelmeyer
Keyword(s):  
Naive Bayes ◽  
Characteristic Curve ◽  
Naïve Bayes ◽  
Real World Data ◽  
Minority Class ◽  
Roc Convex Hull ◽  
Classifier Performance ◽  
Under Sampling ◽  
The Cost ◽  
Normal Class

An approach to the construction of classifiers from imbalanced datasets is described. A dataset is imbalanced if the classification categories are not approximately equally represented. Often real-world data sets are predominately composed of ``normal'' examples with only a small percentage of ``abnormal'' or ``interesting'' examples. It is also the case that the cost of misclassifying an abnormal (interesting) example as a normal example is often much higher than the cost of the reverse error. Under-sampling of the majority (normal) class has been proposed as a good means of increasing the sensitivity of a classifier to the minority class. This paper shows that a combination of our method of over-sampling the minority (abnormal) class and under-sampling the majority (normal) class can achieve better classifier performance (in ROC space) than only under-sampling the majority class. This paper also shows that a combination of our method of over-sampling the minority class and under-sampling the majority class can achieve better classifier performance (in ROC space) than varying the loss ratios in Ripper or class priors in Naive Bayes. Our method of over-sampling the minority class involves creating synthetic minority class examples. Experiments are performed using C4.5, Ripper and a Naive Bayes classifier. The method is evaluated using the area under the Receiver Operating Characteristic curve (AUC) and the ROC convex hull strategy.

A Novel Hybrid-Based Ensemble for Class Imbalance Problem

2018 ◽  
Vol 27 (06) ◽  
pp. 1850025 ◽  
Author(s):  
Huaping Guo ◽  
Jun Zhou ◽  
Chang-an Wu ◽  
Wei She
Keyword(s):  
Class Imbalance ◽  
Imbalanced Data ◽  
Sampling Technique ◽  
Original Training ◽  
Training Set ◽  
Class Imbalance Problem ◽  
Minority Class ◽  
Imbalance Problem ◽  
Under Sampling ◽  
Feature Projection

Class-imbalance is very common in real world. However, conventional advanced methods do not work well on imbalanced data due to imbalanced class distribution. This paper proposes a simple but effective Hybrid-based Ensemble (HE) to deal with two-class imbalanced problem. HE learns a hybrid ensemble using the following two stages: (1) learning several projection matrixes from the rebalanced data obtained by under-sampling the original training set and constructing new training sets by projecting the original training set to different spaces defined by the matrixes, and (2) undersampling several subsets from each new training set and training a model on each subset. Here, feature projection aims to improve the diversity between ensemble members and under-sampling technique is to improve generalization ability of individual members on minority class. Experimental results show that, compared with other state-of-the-art methods, HE shows significantly better performance on measures of AUC, G-mean, F-measure and recall.

Class Imbalance Learning to Heterogeneous Cross Software Projects Defect Prediction

2022 ◽  
Vol 10 (1) ◽  
pp. 0-0
Keyword(s):  
Research Work ◽  
Class Imbalance ◽  
Training Dataset ◽  
Software Projects ◽  
Class Imbalance Problem ◽  
Software Application ◽  
Imbalance Problem ◽  
Under Sampling ◽  
Imbalance Learning ◽  
Class Imbalance Learning

Heterogeneous CPDP (HCPDP) attempts to forecast defects in a software application having insufficient previous defect data. Nonetheless, with a Class Imbalance Problem (CIP) perspective, one should have a clear view of data distribution in the training dataset otherwise the trained model would lead to biased classification results. Class Imbalance Learning (CIL) is the method of achieving an equilibrium ratio between two classes in imbalanced datasets. There are a range of effective solutions to manage CIP such as resampling techniques like Over-Sampling (OS) & Under-Sampling (US) methods. The proposed research work employs Synthetic Minority Oversampling TEchnique (SMOTE) and Random Under Sampling (RUS) technique to handle CIP. In addition to this, the paper proposes a novel four-phase HCPDP model and contrasts the efficiency of basic HCPDP model with CIP and after handling CIP using SMOTE & RUS with three prediction pairs. Results show that training performance with SMOTE is substantially improved but RUS displays variations in relation to HCPDP for all three prediction pairs.

ANALISA PERBANDINGAN DALAM RANGKA PEMILIHAN BEBERAPA IP UPSTREAM PROVIDER DI PT. INDOSAT

2018 ◽  
Vol 4 (1) ◽  
pp. 57
Author(s):  
Yuli Anwar ◽  
Dahlar .
Keyword(s):  
Business Strategy ◽  
High Speed ◽  
Human Life ◽  
Sustainable Growth ◽  
The Internet ◽  
Network Service Provider ◽  
High Speed Data ◽  
Telecommunications Services ◽  
The Cost ◽  
Internet Network

Abstract. One of the advances in information technology that now has changed the outlook and human life, business process and business strategy of an institution is the internet. The internet is a very large networks that connected to computers and serves throughout the world in one centralized network. With the internet we can access data and information anytime and anywhere.    As one provider of high-speed data communications services and the pioneer of the internet network service provider in Indonesia that provides integrated services, as well as one of the pioneer development of internet services that provide extensive services in the building and apply it throughout Indonesia. Indosat ready to seize opportunities for sustainable growth of business spectrum are still sprawling Indonesia.    Therefore, Indosat continues to focus on the development of increased efforts to provide the best service for customers of Indosat. Indosat will continue to develop and expand network coverage and a larger investment that the company will achieve excellence in the field of integrated telecommunications services.    Ranking by region of the IP Providers can be seen by grouping IP Providers, and management over IP Providers prefer to choose providers based on where it orginates as an example for the region of the U.S if it will be preferred providers that come from U.S. providers.With the commencement of the internet network optimization start early in 2008 with the selection of the appropriate IP Upstream Provider criteria, it is up to date according to data obtained from Indosat, seen any significant changes to the cost of purchasing capacity of the IP Upstream.    Based on the data obtained that until Q3 or September 2008, the number of IP Upstream Providers that previously there were 20 to 10 IP Upstream Provider, IP Transit Price total decrease of 11% to the price of IP Transit Price / Mbps there is a decrease of 78%, while from the capacity bandwith an increase of 301% capacity from 2008.

Governing the transboundary risks of offshore methane hydrate exploration in the seabed and ocean floor—an analysis on international provisions and Chinese law†

2020 ◽  
Vol 13 (2) ◽  
pp. 185-203
Author(s):  
Dong Yan ◽  
Paolo Davide Farah ◽  
Tivadar Ötvös ◽  
Ivana Gaskova
Keyword(s):  
Methane Hydrate ◽  
Human Life ◽  
Damage Control ◽  
Legal Framework ◽  
Environmental Harm ◽  
Methane Hydrates ◽  
International Role ◽  
Deep Seabed Mining ◽  
The Cost ◽  
The Unclos

Abstract Considering the fact that its existence is abundant while maintaining the ability to generate freshwater while burning, methane hydrates have been classified as sources of sustainable energy. China currently maintains an international role in developing technology meant to explore offshore methane hydrates buried under the mud of the seabed, their primary laboratory being the South China Sea. However, such a process does not come without its hazards and fatal consequences, ranging from the destruction of the flora and fauna, the general environment, and—the greatest hazard of all—the cost of human life. The United Nations Convention on the Law of the Sea (hereinafter ‘UNCLOS’), being an important international legal regime and instrument, has assigned damage control during the exploration of methane hydrates, as being the responsibilities and liability of individual sovereign states and corporations. China adopted the Deep Seabed Mining Law (hereinafter the DSM Law) on 26 February 2016, which came into force on the 1 of May 2016; a regulation providing the legal framework also for the Chinese government’s role in methane hydrate exploratory activities. This article examines the role of the DSM Law and its provisions, as well as several international documents intended to prevent transboundary environmental harm from arising, as a result of offshore methane hydrate extraction. Despite the obvious risk of harm to the environment, the DSM Law has made great strides in regulating exploratory activities so as to meet the criteria of the UNCLOS. However, this article argues that neither the UNCLOS nor the DSM Law are adequately prepared to address transboundary harm triggered by the exploitation of offshore methane hydrates. In particular, the technology of such extraction is still at an experimental stage, and potential risks remain uncertain—and even untraceable—for cross-jurisdictional claims. The article intends to seek available legal instruments or models, to overhaul the incapacity within the current governing framework, and offers suggestions supporting national and international legislative efforts towards protecting the environment during methane hydrate extraction.

scholarly journals Hybridization of ring theory-based evolutionary algorithm and particle swarm optimization to solve class imbalance problem

2021 ◽  
Author(s):  
Sayan Surya Shaw ◽  
Shameem Ahmed ◽  
Samir Malakar ◽  
Laura Garcia-Hernandez ◽  
Ajith Abraham ◽  
...  
Keyword(s):  
Particle Swarm Optimization ◽  
Real Life ◽  
Class Imbalance ◽  
Ring Theory ◽  
Class Imbalance Problem ◽  
Minority Class ◽  
Swarm Optimization ◽  
Imbalance Problem ◽  
Representative Samples ◽  
Selection Of

AbstractMany real-life datasets are imbalanced in nature, which implies that the number of samples present in one class (minority class) is exceptionally less compared to the number of samples found in the other class (majority class). Hence, if we directly fit these datasets to a standard classifier for training, then it often overlooks the minority class samples while estimating class separating hyperplane(s) and as a result of that it missclassifies the minority class samples. To solve this problem, over the years, many researchers have followed different approaches. However the selection of the true representative samples from the majority class is still considered as an open research problem. A better solution for this problem would be helpful in many applications like fraud detection, disease prediction and text classification. Also, the recent studies show that it needs not only analyzing disproportion between classes, but also other difficulties rooted in the nature of different data and thereby it needs more flexible, self-adaptable, computationally efficient and real-time method for selection of majority class samples without loosing much of important data from it. Keeping this fact in mind, we have proposed a hybrid model constituting Particle Swarm Optimization (PSO), a popular swarm intelligence-based meta-heuristic algorithm, and Ring Theory (RT)-based Evolutionary Algorithm (RTEA), a recently proposed physics-based meta-heuristic algorithm. We have named the algorithm as RT-based PSO or in short RTPSO. RTPSO can select the most representative samples from the majority class as it takes advantage of the efficient exploration and the exploitation phases of its parent algorithms for strengthening the search process. We have used AdaBoost classifier to observe the final classification results of our model. The effectiveness of our proposed method has been evaluated on 15 standard real-life datasets having low to extreme imbalance ratio. The performance of the RTPSO has been compared with PSO, RTEA and other standard undersampling methods. The obtained results demonstrate the superiority of RTPSO over state-of-the-art class imbalance problem-solvers considered here for comparison. The source code of this work is available in https://github.com/Sayansurya/RTPSO_Class_imbalance.

scholarly journals Embedding Undersampling Rotation Forest for Imbalanced Problem

2018 ◽  
Vol 2018 ◽  
pp. 1-15 ◽  
Author(s):  
Huaping Guo ◽  
Xiaoyu Diao ◽  
Hongbing Liu
Keyword(s):  
Imbalanced Data ◽  
Feature Space ◽  
Original Data ◽  
Training Set ◽  
Data Set ◽  
Minority Class ◽  
Rotation Forest ◽  
Novel Method ◽  
Individual Classifier ◽  
The Cost

Rotation Forest is an ensemble learning approach achieving better performance comparing to Bagging and Boosting through building accurate and diverse classifiers using rotated feature space. However, like other conventional classifiers, Rotation Forest does not work well on the imbalanced data which are characterized as having much less examples of one class (minority class) than the other (majority class), and the cost of misclassifying minority class examples is often much more expensive than the contrary cases. This paper proposes a novel method called Embedding Undersampling Rotation Forest (EURF) to handle this problem (1) sampling subsets from the majority class and learning a projection matrix from each subset and (2) obtaining training sets by projecting re-undersampling subsets of the original data set to new spaces defined by the matrices and constructing an individual classifier from each training set. For the first method, undersampling is to force the rotation matrix to better capture the features of the minority class without harming the diversity between individual classifiers. With respect to the second method, the undersampling technique aims to improve the performance of individual classifiers on the minority class. The experimental results show that EURF achieves significantly better performance comparing to other state-of-the-art methods.

scholarly journals IMMERSIVE TECHNOLOGIES AS A TOOL OF INNOVATIVE MARKETING IN THE FIELD OF BUSINESS ORGANIZATION

2021 ◽  
Author(s):  
Valentyna Fostolovych ◽  
Tetiana Botsian
Keyword(s):  
Human Life ◽  
Digital Technologies ◽  
Digital Transformation ◽  
Competitive Advantages ◽  
Marketing Activities ◽  
Marketing Tool ◽  
The Cost ◽  
The Impact ◽  
Immersive Technologies

The permeability of all spheres of both economic activity and human life with digital technologies encourages the search for new marketing ideas necessary for the implementation of the product (goods, works and services).  Today's consumer has become more demanding both to the product itself and to the ways of presenting it.  Immersive technologies are becoming one of the tools that contribute to the formation of competitive advantages, especially the organization of business in the field of entertainment, as one of the areas of additional income in the field of hotel and restaurant services and marketing activities of enterprises.  Digital transformation leads to the search for new initiatives that will be a tool to meet customer needs and a way to reach wider market segments.  The process of digitalization must first be integrated into the economy of the whole state and the enterprise as a whole, and in all processes of production of goods, works and services.  Digital-transformation of domestic enterprises will help to obtain additional competitive advantages both in the domestic market and in the international market.  The formation of competitive advantages is associated not only with the maximum involvement of digital technologies in business.  It is important to choose such technologies that will be most effective in the implementation of a particular type of enterprise, under certain conditions and in a particular environment. The expediency of using immersive technologies as a marketing tool is undeniable.  However, in addition to tools, immersive technologies are important as a means of education, a separate milestone in the field of entertainment, a means of psychological influence and more.  That is, the impact of this tool on the level of competitiveness of the enterprise in the environment of the demanding consumer is manifested: in the form of reducing the cost of attracting the client; active covert promotion through their use; improving the quality of the presented product (goods, works, services); ensuring the elasticity of the enterprise to the needs and requirements of consumers; the transition of the enterprise to an innovative type of development and active digitalization.

scholarly journals An under sampled k-means approach for handlingimbalanced data using diversified distribution

2018 ◽  
Vol 7 (1.8) ◽  
pp. 113 ◽  
Author(s):  
G Shobana ◽  
Bhanu Prakash Battula
Keyword(s):  
Research Issue ◽  
Minority Class ◽  
Effective Utilization ◽  
Specific Strategy ◽  
Open Research ◽  
Novel Approach ◽  
Under Sampling ◽  
F Measure

Some true applications uncover troubles in taking in classifiers from imbalanced information. Albeit a few techniques for enhancing classifiers have been presented, the distinguishing proof of conditions for the effective utilization of the specific strategy is as yet an open research issue. It is likewise worth to think about the idea of imbalanced information, qualities of the minority class dissemination and their impact on arrangement execution. In any case, current investigations on imbalanced information trouble factors have been predominantly finished with manufactured datasets and their decisions are not effortlessly material to this present reality issues, likewise on the grounds that the techniques for their distinguishing proof are not adequately created. In this paper, we recommended a novel approach Under Sampling Utilizing Diversified Distribution (USDD) for explaining the issues of class lopsidedness in genuine datasets by thinking about the systems of recognizable pieces of proof and expulsion of marginal, uncommon and anomalies sub groups utilizing k-implies. USDD utilizes exceptional procedure for recognizable proof of these kinds of cases, which depends on breaking down a class dissemination in a nearby neighborhood of the considered case utilizing k-closest approach. The exploratory outcomes recommend that the proposed USDD approach performs superior to the looked at approach as far as AUC, accuracy, review and f-measure.

Sign in / Sign up

Export Citation Format

Share Document