Fig 2 - uploaded by Moez Krichen
Content may be subject to copyright.
SVM results (F-measure) for each selected number of indicators. The tested date is April 1 st , 2020.
Source publication
The new so-called COVID-19 virus is unfortunately founded to be highly transmissible across the globe. In this study, we propose a novel approach for estimating the spread level of the virus for each country for three different dates between April and May 2020. Unlike previous studies, this investigation does not process any historical data of spre...
Context in source publication
Context 1
... selection configurations are exposed bellow in Table II. As shown in Table II, for April 1st, the best classifier was the SVM with an F-measure equal to 89.06%. This optimal result is obtained with the optimal number of 40 indicators representing the best 40 indicators according to the univariate selection approach. This result is emphasized in Fig. 2 in which we observe increasing scores till 40 selected indicators and then decreasing scores when more indicators were selected (60 indicators and higher numbers). This is explained by the fact that adding irrelevant features may disturb the classification ...
Similar publications
In this study, we aim to estimate the sigma coefficient in the activation probability calculation for a topic’s diffusion prediction problem. In our previous studies, we proposed an aggregated activation probability combination of the metapath and text information, in which sigma is the characteristic coefficient of interest’s similarity based on t...
Sensitivity analysis is a popular feature selection approach employed to identify the important features in a dataset. In sensitivity analysis, each input feature is perturbed one-at-a-time and the response of the machine learning model is examined to determine the feature's rank. Note that the existing perturbation techniques may lead to inaccurat...
Background:
The number of patients with fragility fracture has been increasing. Although the increasing number of patients with fragility fracture increased the rate of fracture (refracture), the causes of refracture are multifactorial, and its predictors are still not clarified. In this issue, we collected a registry-based longitudinal dataset th...
Best group subset selection aims to choose a small part of non-overlapping groups to achieve the best interpretability on the response variable. It is practically attractive for group variable selection; however, due to the computational intractability in high dimensionality setting, it doesn't catch enough attention. To fill the blank of efficient...
In this paper, we try to solve the feature ranking problem through an allocation of information granularity. In many real applications, people are more concerned with an ordered sequence, especially a sequence with a few most important features. However, the outcome of the feature selection methods is often not stable. We proposed an improved granu...
Citations
... O surto do vírus SARS-CoV-2 disseminou-se rápida e globalmente desde o primeiro caso, relatado no final de dezembro de 2019, culminando na declaração de uma pandemia global pela Organização Mundial da Saúde (OMS) em 11 de março de 2020 (Mihoub et al., 2020). A COVID-19 impactou a saúde humana, registrando mais de 270 milhões de casos e 5,3 milhões de mortes até dezembro de 2021. ...
A pandemia da COVID-19, desencadeada pelo novo coronavírus (SARS-CoV-2), teve início em dezembro de 2019 na China, gerando grande preocupação com a saúde global. Este artigo busca revelar as relações diretas e indiretas entre indicadores socioeconômicos e ambientais e os dados relativos à pandemia da COVID-19, por meio do emprego de técnicas de aprendizado de máquina. Inicialmente, foram identificados os estudos mais relevantes publicados durante o período pandêmico e pós-pandêmico que abordam o tema proposto. Após uma triagem completa, a pesquisa destacou as principais referências da literatura. Finalmente, este trabalho utilizou técnicas de descoberta de conhecimento em bases de dados da COVID-19 e de indicadores socioeconômicos e ambientais. Para abordar o problema em estudo, o método empregado consistiu nas etapas de seleção, pré-processamento, transformação, mineração de dados e avaliação dos resultados. Resultados significativos foram obtidos com a aplicação de métodos de mineração de dados descritiva, envolvendo modelos de correlação, agrupamento e regras de associação; estas técnicas se destacaram, apresentando capacidades de generalização satisfatórias. Os resultados desta pesquisa têm o potencial de contribuir para a formulação de políticas públicas e para a tomada de decisões futuras relacionadas à saúde pública.
... Mitchell provides a classic but serviceable definition in [8]. Researchers are interested in ML because it can help them answer important scientific and technical questions, and because it can be used to enhance the highly useful computer software that is produced and used in many applications, including the medical field [11]- [15]. In this research, we suggest an optimized and hybridized machine learning approach to automatic Dyslexia detection. ...
Dyslexia is a form of specialized learning disability that can lead to academic dissatisfaction. Detection is essential yet difficult, especially for languages like Arabic and Spanish that have straightforward orthographies. As seen by the abundance of work given in a wide range of scientific journals, the detection of dyslexia continues to be a key pursuit for the research efforts in a variety of domains. It appears that machine learning and other statistical approaches are quite useful for these tasks, such as those used in the classification of medical datasets. In this paper, we combined principle machine learning models for dyslexia detection. However, a lot of work remains to be done by hand in order to properly categorize the enormous dataset employed during training. Recent research have shown that machine learning models are effective, proving their ability to lessen the workload associated with feature engineering. However, applying the machine learning model directly to the classification task yields no statistically significant gain in performance. For better machine learning prediction, we introduce an optimal ensemble learning model based on a genetic algorithms. Dyslexia feature extraction is first, followed by a hybridization of algorithms for machine learning in step tow of the model. After that, we use an adaptive genetic algorithm for select the most optimal weight to combine different machine learning models. We conducted trials on the Dyslexia dataset to show that our proposed technique significantly increases accuracy.
... The kernel function that represents the second essential concept provides the answer by translating the original data space into a higherdimensional space, where it is probable to obtain a linear separator. The SVM concept handles nonlinear classification problems effectively in this manner [27,28]. An illustration of the SVM Algorithm is shown in Figure 8. ...
The state of roads may sometimes be difficult to perceive due to intense climate conditions, absence of road signs, or simply human inattention, which may be harmful to both vehicles and drivers. The automatic monitoring of the road states represents a promising solution to warn drivers about the status of a road in order to protect them from injuries or accidents. In this paper, we present a novel application for data collection regarding road states. Our application entitled “Road Scanner” allows onboard users to tag four types of segments in roads: smooth, bumps, potholes, and others. For each tagged segment the application records multimodal data using the embedded sensors of a smartphone. The collected data concerns mainly vehicle accelerations, angular rotations, and geographical positions recorded by the accelerometer, the gyroscope, and the GPS sensor, respectively, of a user phone. Moreover, a medium-size dataset was built and machine learning models were applied to detect the right label for the road segment. Overall, the results were very promising since the SVM classifier (Support Vector Machines) has recorded an accuracy rate of 88.05%.
... As stated earlier, hateful and offensive speech has grown significantly on social media in the last couple of years, especially in the aftermath of the COVID-19 lockdown. With the rise of Artificial Intelligence (AI) methods and their great performance in many fields [6][7][8][9][10], several researchers from around the globe have also presented several AI-based solutions to automate the detection of hateful and offensive speech in their online communities. ...
The democratization of access to internet and social media has given an opportunity for every individual to openly express his or her ideas and feelings. Unfortunately, this has also created room for extremist, racist, misogynist, and offensive opinions expressed either as articles, posts, or comments. While controlling offensive speech in English-, Spanish-, and French- speaking social media communities and websites has reached a mature level, it is much less the case for their counterparts in Arabic-speaking countries. This paper presents a transfer learning solution to detect hateful and offensive speech on Arabic websites and social media platforms. This paper will compare the performance of different BERT-based models trained to classify comments as either abusive or neutral. The training dataset contains comments in standard Arabic as well as four dialects. We will also use their English translations for comparative purposes. The models were evaluated based on five metrics: Accuracy, Precision, Recall, F1-Score, and Confusion Matrix.
... In order to reduce the effect of yearly fluctuations in the data, we have taken moving averages of three years (i.e., 2018 to 2020 for the food security indicators). Moreover, we used the moving averages of three years because there were several missing values in the yearly data, a similar approach was employed in the study of Tamasiga et al., [23] who investigated the impact of socio-economic indicators on COVID-19 and Mihoub et al., [12] who applied machine learning techniques in order to make predictions of cases of COVID-19. ...
The outbreak of COVID-19 led to the implementation of lockdowns and social distancing regulations to curb the spread of infections. Consequently, the lockdowns impeded the movement of smallholder farmers, agricultural inputs, and food products thereby disrupting the food supply chains in SSA. Therefore, this paper examines the relationship between food security indicators (accessibility, availability, utilization, stability) and COVID-19. This study uses ordinary least square regression (OLS) models to study the relationship between the food security indicators and COVID-19. The study considers 9 out of 48 sub-Saharan African countries (Benin, Burkina Faso, Cameroon, Chad, Madagascar, Mali, Mauritania, Nigeria, Senegal) due to data availability restrictions. The result of the analysis indicated that a rise in COVID-19 levels negatively impacts all the 4 indicators of food security without exception. This paper underscores the need to consider the disruptions of food security indicators such as diet, nutritional content, access and availability , affordability, and food supply chains. Moreover, the paper discusses mitigating strategies that may alleviate SSA's food security amidst the COVID-19 pandemic. We recommend that SSA countries invest in quality agricultural and food production infrastructure and supporting industries that contribute directly to the food supply chain, such as agro-processing, fertilizer production and transport. Another important dimension of the COVID-19 and food insecurity syndemic is the income shocks that occurred as a consequence of the COVID-19 outbreak. Like many factories, companies, and service providers closed shop (especially the informal sector), people lost their incomes as a result of loss of employment and, in many instances, no social protection. Therefore, we recommend that SSA governments develop affordable, sustainable, and targeted social protection/insurance systems that extend to the informal sector of the economy.
... AI has an incredible impact on different socioeconomic fields, and it is expected by 2030 to add up to $13 trillion to the GDP [31]. AI has demonstrated its capacities in different fields [32][33][34][35][36]. For instance, the authors in [37] suggest a sparse regression-based, data-driven model identification for ships with 6 degrees of freedom. ...
River floods are listed among the natural disasters that can directly influence different aspects of life, ranging from human lives, to economy, infrastructure, agriculture, etc. Organizations are investing heavily in research to find more efficient approaches to prevent them. The Artificial Intelligence of Things (AIoT) is a recent concept that combines the best of both Artificial Intelligence and Internet of Things, and has already demonstrated its capabilities in different fields. In this paper, we introduce an AIoT architecture where river flood sensors, in each region, can transmit their data via the LoRaWAN to their closest local broadcast center. The latter will relay the collected data via 4G/5G to a centralized cloud server that will analyze the data, predict the status of the rivers countrywide using an efficient Artificial Intelligence approach, and thus, help prevent eventual floods. This approach has proven its efficiency at every level. On the one hand, the LoRaWAN-based communication between sensor nodes and broadcast centers has provided a lower energy consumption and a wider range. On the other hand, the Artificial Intelligence-based data analysis has provided better river flood predictions.
... Previous studies have demonstrated that the effective size for random forest is varied [14]- [16] and there is no exact number or formula found yet that determines the effective sample size. Referring to [15] that applies random forest to the same database (e.g., World Bank Development Indicators) to classify and predict, we set a limit on number of host parties more than 12, which leaves ve projects for random forest including Biogas, Biomass, Hydro Power, Methane recovery & utilization and Wind Power. ...
The performance of climate technology implemented under Clean Development Mechanism (CDM) is varied in terms of Certified Emission Reductions (CERs) issuance rate, a ratio of the actual CER issued of the total CER expected, ranged between 2% and 193%. In this study, we propose a novel approach for estimating the CER issuance rate (i.e., lower or higher than 100%) by using random forest algorithms. Unlike previous studies, we associate 1443 indicators related to economy, environment and society from 28 countries with the outcome of CER issuance rate of 3338 CDM projects and examine mixed effects of local context-driven indicators. Data preprocessing and feature selection techniques are applied to extract important indicators with which classification and prediction are conducted across different types of CDM technologies. The highest accuracy and F-measure of the developed prediction model are achieved from wind projects (84%, 92%, respectively). The obtained results exhibit possibilities of capturing the performance gap through machine learning algorithm. Moreover, we recommend adopting local demand-driven indicators to enhance the usability of the proposed approach.
... Thousands of diagnostic factors linked to healthcare difficulties can have an impact on the severity of symptoms. As a result, advanced feature selection approaches are trained to determine those attributes that contribute more to the condition of the patient we are concerned with, as possessing unnecessary clinical manifestations considerably raises the possibilities of forecasting inaccuracy [6]. There are numerous features extraction methodologies. ...
... It has 27 layers (including the pooling layers) and employs average pooling. In order to address the problem of fading gradients, the story of Inception makes use of two Begin Set the default values for Convolution Max pooling Layer in [1][2][3] for Fully Connected Layer in [1][2][3] for Number of Filter in [16,18,20,22,24] for Filter size in [3][4][5][6] model = D-CNN_train score =D-CNN_forecast return max Algorithm 1: Proposed algorithm. 8 BioMed Research International additional classifications. ...
As an epidemic, COVID-19's core test instrument still has serious flaws. To improve the present condition, all capabilities and tools available in this field are being used to combat the pandemic. Because of the contagious characteristics of the unique coronavirus (COVID-19) infection, an overwhelming comparison with patients queues up for pulmonary X-rays, overloading physicians and radiology and significantly impacting the quality of care, diagnosis, and outbreak prevention. Given the scarcity of clinical services such as intensive care and motorized ventilation systems in the aspect of this vastly transmissible ailment, it is critical to categorize patients as per their risk categories. This research describes a novel use of the deep convolutional neural network (CNN) technique to COVID-19 illness assessment seriousness. Utilizing chest X-ray images as contribution, an unsupervised DCNN model is constructed and suggested to split COVID-19 individuals into four seriousness classrooms: low, medium, serious, and crucial with an accuracy level of 96 percent. The efficiency of the DCNN model developed with the proposed methodology is demonstrated by empirical findings on a suitably huge sum of chest X-ray scans. To the evidence relating, it is the first COVID-19 disease incidence evaluation research with four different phases, to use a reasonably high number of X-ray images dataset and a DCNN with nearly all hyperparameters dynamically adjusted by the variable selection optimization task.
... There are two strands of literature studying the correlation of COVID-19 cases and socio-economic indicators. The first strand is the literature focusing on the spread of COVID-19, like those of Mihoub et al. (2020). The second strand of literature describes correlations between socio-economic status, demographic factors and COVID-19. ...
... Due to many missing values for 2019 and 2020, the 2018 data were extracted. A similar approach was adopted in a study by Mihoub et al. (2020) that applied machine learning techniques to predict COVID-19 cases in a multi-country model. Due to many missing indicators, Mihoub et al. (2020) implemented a machine learning model on COVID 2020 data using socio-economic indicators from 2017. ...
... A similar approach was adopted in a study by Mihoub et al. (2020) that applied machine learning techniques to predict COVID-19 cases in a multi-country model. Due to many missing indicators, Mihoub et al. (2020) implemented a machine learning model on COVID 2020 data using socio-economic indicators from 2017. The stages of data preparation to running the model and obtaining the results are summarised below: ...
The COVID-19 pandemic has triggered an unprecedented social and economic crisis. This study aims at investigating the impact of socio-economic indicators on the levels of COVID-19 (confirmed and death cases) in sub-Saharan Africa. The investigation makes use of the readily accessible public data: we obtain COVID-19 data from Johns Hopkins and socio-economic indicators from the World Bank. The socio-economic indicators (independent variables) used in the multilinear regression were GDP per capita, gross national income per capita, life expectancy, population density (people per sq. km of land area), the population aged 65 and above, current health expenditure per capita and total population. The dependent variables used were the COVID-19 confirmed and death cases. Amongst the seven socio-economic indicators, only 4 showed a statistically significant impact on COVID-19 cases: population density, gross national income per capita, population aged 65 and above and total population. The obtained R2 of 69% and 63% indicated that the socio-economic indicators captured and explained the variation of COVID-19 confirmed cases and COVID-19 death cases, respectively. The startling results obtained in this study were the negative but statistically significant relationship between COVID-19 deaths and population density and the positive and statistically significant relationship between gross national income per capita and COVID-19 cases (both confirmed and deaths). Both these results are at odds with literature investigating these indicators in Europe, China, India and the UK.
... O surto do vírus SARS-CoV-2 se espalhou rapidamente por todo o mundo após o primeiro caso relatado no final de dezembro de 2019 e foi declarada uma pandemia global pela Organização Mundial da Saúde -OMS (Organization World Health -WHO) em 11 março de 2020 [Mihoub et al. 2020]. A COVID-19 afetou a saúde humana como também a economia global. ...
... Em seguida, serão apresentados de forma breve os trabalhos que obtiveram maior relação com as questões de pesquisa elencadas nesta RSL. [Mihoub et al. 2020] propõem estudar a correlação entre a propagaç˜do ao COVID-19 e os indicadores socioeconômicos de diferentes países usando técnicas de inteligência artificial. Os níveis de propagação foram divididos em três datas diferentes, entre abril e maio de 2020. ...
A pandemia da COVID-19, causada pelo novo coronavírus (SARS-CoV-2), iniciou-se em dezembro de 2019 na China, colocando a saúde da população mundial em grande alerta. Este artigo objetiva apresentar uma Revisão Sistemática da Literatura que análise a correlação entre indicadores socioeconômicos e a pandemia da COVID-19 através de técnicas de aprendizagem de máquina, identificando quais técnicas e software de análise de dados foram utilizados. Foram identificados os estudos mais relevantes publicados no período de 2018 a 2022, que abordam o tema proposto. Após a triagem completa a pesquisa destacou 10 artigos selecionados no total de 814 analisados.
