Context & scaleAlthough most people born this century will be educated in African schools, these schools often lack basic infrastructure, such as electricity and/or lighting. In the face of a rapidly growing school-age population Africa, electrification educational facilities is not just an infrastructural challenge but also pivotal investment continent's future workforce. This study reveals stark reality: third Africa's school-aged children are nearer to without electricity, impacting...

There has been a rapid evolution of tree-based ensemble algorithms which have outperformed deep learning in several studies, thus emerging as competitive solution for many applications. In this study, ten (random forest, bagging meta-estimator, adaptive boosting (AdaBoost), gradient machine (GBM), extreme (XGBoost), light (LightGBM), histogram-based GBM, categorical (CatBoost), natural (NGBoost), and the regularised greedy forest (RGF)) were comparatively evaluated enhancement Copernicus...

The growth of Internet Things (IoT) brings the promise a wide range new recommender systems due to expected 57 billion smart connected devices by 2025. In this paper, we propose IoT platform for supporting real-time system. To illustrate effectiveness our proposed platform, present prototype implementation and tourism application demonstrate entire process from user event data collection notification/recommendations provision. We conducted several experiments including notification system...

Abstract Component fault detection and inventory are one of the most significant bottlenecks facing electricity transmission distribution utility establishments especially in developing countries for delivery efficient services to customers ensure proper asset audit management network optimization load forecasting. For lack technology data, insecurity, complexity associated with traditional methods, untimeliness, general human cost, assets monitoring, have remained a big problem many...

Abstract. The correction of digital elevation models (DEMs) can be achieved using a variety techniques. Machine learning and statistical methods are broadly applicable to DEM case studies in different landscapes. However, literature survey did not reveal any research that compared the effectiveness or performance both methods. In this study, we comparatively evaluate three gradient boosted decision trees (XGBoost, LightGBM CatBoost) multiple linear regression for two publicly available...

Modelling topological relationships between places and events is challenging especially because these are dynamic, their evolutionary analysis relies on the explanatory power of representing interactions across different temporal resolutions. In this paper, we introduce Space-Time Varying Graph (STVG) based whole graph approach that combines directed bipartite subgraphs with a time-tree for complex interaction time. We demonstrate how proposed STVG can be exploited to identify extract...

Sequence of graph snapshots have been commonly utilized in literature to represent changes a dynamic graph. This approach may be suitable for small-size and slowly evolving graphs; however, it is associated with high storage overhead massive fast-evolving graphs because replication the entire from one snapshot another at shorter temporal resolutions. presents drawback especially where efficient evolutionary analytics relies on explanatory power representing dynamics across different In this...

Computer vision for large scale building detection can be very challenging in many environments and settings even with recent advances deep learning technologies. Even more is modeling to detect the presence of specific buildings (in this case schools) satellite imagery at a global scale. However, despite variation school structures from rural urban areas country country, have identifiable overhead signatures that make them possible detected high-resolution modern techniques. Our hypothesis...

Understanding the biases in Deep Neural Networks (DNN) based algorithms is gaining paramount importance due to its increased applications on many real-world problems. A known problem of DNN penalizing underrepresented population could undermine efficacy development projects dependent data produced using DNN-based models. In spite this, problems for Land Use and Cover Classification (LULCC) have not been a subject studies. this study, we explore ways quantify land use with an example...

GPS-equipped public transit vehicles generate a massive amount of location information, yet analytical methods based on Geographic Information System and Relational Database Management Systems are limited in their ability to handle these data for performance assessment. Graph analytics approach appears well suited addressing limitations; however, existing graph models that have been used represent the network do not provide flexibility incorporate mobility context from Automatic Vehicle...

Land Surface Temperature (LST) is one of the factors associated to urban heat rise and micro climatic warming within a city. Researches relating development new technologies or improvement on existing ones are very important in climate studies. This paper expounds our study simulation prediction specific future time LST quantitative trend Ikom city Nigeria using Feed Forward Back Propagation Artificial Neural Network technology. was based series ANN model that takes sequence past values,...

Several methods have been proposed for correcting the elevation bias in digital models (DEMs) example, linear regression. Nowadays, supervised machine learning enables modelling of complex relationships between variables, and has deployed by researchers a variety fields. In existing literature, several studies adopted either or statistical approaches task DEM correction. However, to our knowledge, none these compared performance both approaches, especially with regard open-access global...

Abstract. Gradient-Boosted Decision Trees (GBDTs), particularly when tuned with Bayesian optimisation, are powerful machine learning techniques known for their effectiveness in handling complex, non-linear data. However, the performance of these models can be significantly influenced by characteristics terrain being analysed. In this study, we assess three Bayesian-optimised GBDTs (XGBoost, LightGBM and CatBoost) using digital elevation model (DEM) error correction as a case study. The is...

Mobility analytics using data generated from the Internet of Mobile Things (IoMT) is facing many challenges which range ingestion streams coming a vast number fog nodes and IoMT devices to avoiding overflowing cloud with useless massive that can trigger bottlenecks [1]. Managing flow becoming an important part because it will dictate in platform analytical tasks should run future. Data flows are usually sequence out-of-order tuples high input rate, mobility requires real-time both...

The study of the dynamic relationship between topological structure a transit network and mobility patterns vehicles on this is critical towardsdevising smart time-aware solutions to management recommendation systems. This paper proposes time-varying graph (TVG) model thisrelationship. effectiveness proposed has been explored by implementing in Neo4j database using feeds generated bus City Moncton, New Brunswick, Canada. Dynamics relationshipalsohave detected metrics such as temporal...

Environmental monitoring and management systems in most cases deal with models spatial analytics that involve the integration of in-situ remote sensor observations. In-situ observations those gathered by sensors are usually provided different databases services real-time dynamic such as Geo-Web Services. Thus, data have to be pulled from transferred over network before they fused processed on service middleware. This process is very massive unnecessary communication work load service....

Abstract Fault identification is one of the most significant bottlenecks faced by electricity transmission and distribution utilities in developing countries to deliver efficient services customers ensure proper asset audit management for network optimization load forecasting. This due data scarcity, inaccessibility insecurity, ground-surveys complexity, untimeliness, general human cost. In view this, we exploited use oblique UAV imagery with a high spatial resolution fine-tuned deep...

The interplay between Geographical Information System (GIS) and Computer Science has continued to yield improved methods of carrying out many surveying-related activities. In the past, survey control points were stored in file systems at best Database Management applications thereby leading limited usage since they are difficult locate field. This study however, suggests another approach for storage these which makes them be easily accessible gives room faster update geo-visualization...

Abstract Graph‐pattern association rules have been explored for detecting frequent subgraph structures in real‐world network data, which can reveal new insights decision‐making, recommender systems, and predictive models. However, questionnaire data neglected so far even though they are one of the most affordable ways to gather quantitative data. Questionnaires cover every aspect a topic, generating strategies trends many organisations. The challenge is twofold: develop model handling...

A significant outcome of the COVID-19 pandemic has been hyper dependency on digital tools. As schools and, at many times, businesses have converted online, use tools increased rapidly. Since situation fluctuates, national and state-level regulations caused to switch back forth between traditional online formats. When deciding two, governments should factor in effects mental health prevent rates depression. Our research aims analyze correlation depression implement an autoregressive model...