2023 - Repozytorium publikacji - Politechnika Gdańska

Pokaż publikacje z roku

Publikacje z roku 2023

Pokaż wszystkie

Universities’ journey towards sustainability- systematic literature review
- Magdalena Popowska
- Monika Sady
2023 Pełny tekst International Journal of Sustainability in Higher Education
Purpose: This research aims to identify a sustainable university's key features. It is an essential step in tracing the topics discussed in the context of a sustainable university and their evolution in the scientific discourse. Design/methodology/approach: This paper relies on a systematic literature review (SLR) conducted using two scholarly databases: Emerald and Scopus. The timeframe selected by the authors for reviewing the available sources spans from 2001 to 2021. Findings: The analysis distinguished seven sustainable university categories, each revealing critical features of sustainable higher education. Each of these categories represents an intriguing area for in-depth analysis. The systematic literature review (SLR) reveals gaps requiring further scientific exploration. Originality: SLR is a cornerstone of research synthesis and helps integrate scientific evidence from qualitative and quantitative published studies. Conducted research presents knowledge about university sustainability and can help scientists find research gaps. Research limitations/implications: The performed literature review was determined by the choice of entries (keywords) to identify the scientific papers in the selected databases. Moreover, since the authors aimed to focus on peer-reviewed sources, this SLR did not include books and doctoral dissertations dealing with the studied issues. The results of the analysis can be used practically by both researchers and practitioners in the field of sustainable development. Identified scientific gaps become a potential research field, and practitioners interested in the transition toward sustainable development (SD) may contribute by accompanying universities in this journey. Collaboration and networking with business stakeholders are critical vectors for spreading the idea of SD. Society's growing concern for climate change requires accurate and specific actions from institutions. As entities educating future generations, universities have a unique role in transforming toward SD. Our findings allow us to get acquainted with the existing main activities undertaken by HEIs in this field and understand the importance of this topic for researchers.
Unlocking the Secrets of River Pollution: Analyzing Organic Pollutants in Sediments—Experimental Study
- Sadeq Abdullah Abdo Alkhadher
- Suhaimi Suratman
- Hussein Al-Hazmi
- Mohamad Pauzi Zakaria
- Bartosz Szeląg
- Joanna Majtacz
- Jakub Drewnowski
2023 Pełny tekst Water
Untreated wastewater released into rivers can result in water pollution, the spread of waterborne diseases, harm to ecosystems, contamination of soil and groundwater, as well as air pollution and respiratory problems for nearby humans and animals due to the release of greenhouse gases. The current study aims to investigate the recent input of anthropogenic loads into the rivers using linear alkylbenzene (LAB), which is one of the molecular chemical markers with application of sophisticated model statistical analyses. In order to determine the compositions of LABs, which act as wastewater pollution molecular indicators, surface sediment samples from the Muar and Kim Kim rivers were collected. Gas chromatography-mass spectrometry (GC-MS) was utilized to identify LABs and investigate their sources and degradation. ANOVA and the Pearson correlation coefficient were employed to determine the significance of differences between sampling locations, with a threshold of p < 0.05. To assess the degradation degree and efficacy of wastewater treatment plants (WWTPs), LABs were identified based on chains ranging from long to short (L/S), C13/C12 homolog, and internal to external (I/E) congeners. The results indicated that LAB concentrations in the studied areas of the Muar River ranged from 87.4 to 188.1 ng g−1dw. There were significant differences in LAB homology at p < 0.05, and a significant percentage of sampling stations contained C13-LAB homology. Based on the LAB ratios (I/E) determined, which ranged from 1.7 to 2.2 in the studied areas, it was concluded that effluents from primary and secondary sources are being discharged into the marine ecosystem in those areas. The degradation of LABs was up to 43% in the interrogated locations. It can be inferred that there is a requirement for enhancing the WWTPs, while also acknowledging the efficacy of LAB molecular markers in identifying anthropogenic wastewater contamination.
Unprecedented Adhesive Performance of Propylene-Based Hydroxyl Functionalized Terpolymers
- Jakub Kruszyński
- Weronika Nowicka
- Miloud Bouyahyi
- Yingxin Liu
- Lanti Yang
- Artur Rozanski
- Nithish Anbuchezhian
- Lidia Jasińska-Walc
- Rob Duchateau
2023 ACS Applied Polymer Materials
The synthesis of hydroxyl functionalized propylene-based terpolymers and their performance as hot melt adhesives were investigated. The products comprise uniformly distributed butyl and 4-hydroxyl-butyl branches along the polypropylene backbone. Despite the low hydroxyl-functionality level of ≤ 0.5 mol%, hydroxyl functionalized terpolymers show formidable adhesion to aluminum and steel providing an adhesive strength exceeding 16 MPa, whereas the non-functionalized congeners hardly adhere to these metals. As evidenced by rheological measurements, the functional groups form dynamic cross-links based on hydrogen bonding and electrostatic interactions with aluminum oxide hydroxide residues, remaining in the product after polymerization. At the industrial application temperature of 180 °C, non-deashed and deashed samples of polymers having 0.1 mol% or 0.5 mol% of incorporated 5-hexen-1-ol gave, upon cooling to room temperature, comparable adhesive strengths. Deashing and increasing the functionality level leads to a significant improvement of the adhesion strength at lower application temperature (130 °C), allowing application of the hydroxyl functionalized propylene-based terpolymers as high strength hot melt adhesive for combinations of polypropylene and metals.
Unraveling a novel microwave strategy to fabricate exposed {001}/{101} facets anatase nanocrystals: Potential for use to the elimination of environmentally toxic metronidazole waste
- Adam Kubiak
- Anna Grzegórska
- Elżbieta Gabała
- Joanna Zembrzuska
- Anna Zielińska-Jurek
- Michał Cegłowski
2023 MATERIALS RESEARCH BULLETIN
This study present a novel microwave strategy to fabricate highly active anatase particles, exposing {101} and {001} facets. Microwave treatment time was shown to determine the growth of crystals in a certain direction. To the best of our knowledge, it is the first report revealing that the contact time of TiO2 crystals with fluorine ions during the microwave process affects the formed morphology, in particular exposed facets ratio. The important element of this study was a parallel approach to produce active photocatalysts and a convenient UV-LED light source. The simultaneous use of synthesized anatase nanocrystals with exposed {001}/{101} facets and tailormade LED photoreactor led to the enhanced elimination of metronidazole. Photocatalytic studies showed greater activity in the elimination of metronidazole with increasing microwave exposure time, which was associated with the presence of the surface heterojunction {001}/{101}, leading to better separation of charge carriers.
Unraveling the Interplay between DNA and Proteins: A Computational Exploration of Sequence and Structure-Specific Recognition Mechanisms
- Kazi Hossain
2023 Pełny tekst
My PhD dissertation focused on DNA-protein interactions and the recognition of specific DNA sequences and structures. I discovered that acidic amino acid residues (Asp/Glu) play a crucial role by exhibiting a preference for cytosine. Their contribution to binding affinity depends on nearby cytosines, balancing electrostatic repulsion with specific interactions. Acidic residues act as negative selectors, discouraging non-cytosine binding, but can be favorable with increasing proximal cytosine count. They exclusively recognize cytosine due to electrostatic repulsion with adenine's N7 atom and stronger hydrogen bonding. In another aspect of my research, I explored conformation-specific DNA recognition. I found that the EXOG protein prefers A-DNA and selectively recognizes RNA/DNA chimeric duplexes. Specific arginine residues induce the A-DNA conformation when EXOG binds to DNA/DNA duplexes, providing insights into mitochondrial replication and base excision repair. Furthermore, I investigated the DHX36 helicase, which recognizes G-quadruplexes (G4s) through its DSM and OB subdomains. The planar face of a G-tetrad and the specific backbone conformation of a G-tract are critical features in this interaction. The DSM and OB subdomains cooperatively recognize these distinctive features of parallel G4s. Importantly, the recognition by DSM is mediated through van der Waals contacts and hydrophobic interactions, exhibiting a preference for the accessible 5'-side of the G4 structure.
UNRES-GPU for Physics-Based Coarse-Grained Simulations of Protein Systems at Biological Time- and Size-Scales
- Krzysztof Ocetkiewicz
- Cezary Czaplewski
- Henryk Krawczyk
- Agnieszka Lipska
- Józef Liwo
- Jerzy Proficz
- Adam Sieradzan
- Paweł Czarnul
2023 Pełny tekst BIOINFORMATICS
The dynamics of the virus like particles (VLPs) corresponding to the GII.4 Houston, GII.2 SMV, and GI.1 Norwalk strains of human noroviruses (HuNoV) that cause gastroenteritis was investigated by means of long-time (about 30 μs in the laboratory timescale) molecular dynamics simulations with the coarse-grained UNRES force field. The main motion of VLP units turned out to be the bending at the junction between the P1 subdomain (that sits in the VLP shell) and the P2 subdomain (that protrudes outside) of the major VP1 protein, this resulting in a correlated wagging motion of the P2 subdomains with respect to the VLP surface. The fluctuations of the P2 subdomain were found to be more pronounced and the P2 domain made a greater angle with the normal to the VLP surface for the GII.2 strain, which could explain the inability of this strain to bind the histo-blood group antigens (HBGAs).
Unveiling the Green Mirage: Unearthing Weaknesses in Pro-Environmental Applications
- Ewa Duda
- Hanna Obracht-prondzyńska
- Helena Anacka
- Jolanta Kowal
2023 Pełny tekst
Growing ecological concerns and a general shift towards tailored digitized solutions in cities lead to an increased popularity of pro ecological applications, especially those associated with transport-oriented functionalities. This research focuses on the results from a testbed of a pro-environmental application deployment among early adopters. Based on quantitative and cartographic analysis enriched with interviews’ results, by applying a triangulation research approach, we are able to answer three research questions: related to potential constraints or unintended negative outcomes of pro-environmental educational applications, deviation from the intended use of these apps, as well as range of climate- responsive behaviors these applications promote. Research results allow us to analyze to what extent a pro environmental application affects users’ behavior and ecological choices. It also helps to critically evaluate potential technological weaknesses, their interaction with end users and an impact on motivations and habits of urban dwellers.
Unveiling the Pool of Metallophores in Native Environments and Correlation with Their Potential Producers
- Francisco Calderón Celis
- Ivan González-Álvarez
- Magdalena Fabjanowicz
- Simon Godin
- Laurent Ouerdane
- Béatrice Lauga
- Ryszard Łobiński
2023 ENVIRONMENTAL SCIENCE & TECHNOLOGY
For many organisms, metallophores are essential biogenic ligands that ensure metal scavenging and acquisition from their environment. Their identification is challenging in highly organic matter rich environments like peatlands due to low solubilization and metal scarcity and high matrix complexity. In contrast to common approaches based on sample modification by spiking of metal isotope tags, we have developed a two-dimensional (2D) Solid-phase extraction-Liquid chromatography-mass spectrometry (SPE-LC-MS) approach for the highly sensitive (LOD 40 fmol per g of soil), high-resolution direct detection and identification of metallophores in both their noncomplexed (apo) and metal-complexed forms in native environments. The characterization of peat collected in the Bernadouze (France) peatland resulted in the identification of 53 metallophores by a database mass-based search, 36 among which are bacterial. Furthermore, the detection of the characteristic (natural) metal isotope patterns in MS resulted in the detection of both Fe and Cu potential complexes. A taxonomic-based inference method was implemented based on literature and public database (antiSMASH database version 3.0) searches, enabling to associate over 40% of the identified bacterial metallophores with potential producers. In some cases, low completeness with the MIBiG reference BCG might be indicative of alternative producers in the ecosystem. Thus, coupling of metallophore detection and producers’ inference could pave a new way to investigate poorly documented environment searching for new metallophores and their producers yet unknown.
Upgrading low value-added activities in global value chains: a functional specialisation approach
- Aleksandra Kordalska
- Magdalena Olczyk
2023 Pełny tekst Economic Systems Research
This paper aims to identify patterns of functional specialisation (FS) in global value chains (GVCs) and determinants of upgrading them for selected Central Eastern European (CEE) economies. By combing the World Input-Output Database with data on occupations, we reveal a new FS pattern among subgroups of CEEs. Poland and Slovakia have an unfavourable GVC position and specialise in low value-added fabrication function. In contrast, other CEEs have competitive advantages in high value-added tasks: the Baltic countries and Slovenia in management services, the Czech Republic and Slovenia in R&D. We identify upgrading factors for different types of FS in GVCs. The wages convergence of CEEs with developed economies, and strong GVC backward linkages support the path to higher value-added in almost all business functions. Higher GDP per capita and lower economic distance to Germany allow CEEs to escape from ‘factory economies' status and also generate higher value-added in R&D activities.
Upoważnienie do zaciągania zobowiązań na rzecz jednostek samorządu terytorialnego jako naruszenie dyscypliny finansów publicznych
- Alicja Sekuła
- Grzegorz Garbacz
2023 Pełny tekst ZESZYTY PRAWNICZE BIURA ANALIZ SEJMOWYCH
Artykuł poświęcony jest zagadnieniu upoważnienia do zaciągania zobowiązań na rzecz jednostek samorządu terytorialnego. W teoretycznej części zdefiniowano dwa pojęcia: zobowiązań oraz zaciągania zobowiązań. Następnie scharakteryzowano zagadnienie podmiotowego i przedmiotowego zaciągania zobowiązań. Nieprzestrzeganie przepisów dotyczących zaciągania zobowiązań, które zawarte są głównie w ustawie o finansach publicznych, prowadzi do naruszenia dyscypliny finansów publicznych. Problematykę tę przeanalizowano z punktu widzenia naruszenia dyscypliny finansów publicznych w związku z zaciąganiem zobowiązań. W części empirycznej artykułu omówiono trzy studia przypadków, tzn. opisano wyroki Głównej Komisji Orzekającej w sprawach o naruszenie dyscypliny finansów publicznych, dotyczące niezgodnego z prawem zaciągania zobowiązań.
Uproszczone metody obliczania wyboczenia silosów z blachy falistej wzmocnionych słupami na podstawie analizy MES i wyników doświadczalnych
- Karol Rejowski
2023 Pełny tekst
Rozprawa doktorska dotyczy zagadnienia stateczności i nośności granicznej stalowych silosów z płaszczem z blachy falistej wzmocnionych pionowymi słupami. Przeprowadzono zaawansowane analizy numeryczne MES oraz badania doświadczalne dotyczące wyboczenia/stateczności analizowanych silosów. Analizy numeryczne zostały przeprowadzone w zakresie liniowej oraz nieliniowej analizy stateczności, z uwzględnieniem wstępnych imperfekcji geometrycznych w kształcie powszechnie stosowanej formy wyboczeniowej LBA oraz w kształcie rzeczywistych imperfekcji pomierzonych geodezyjnie dla rzeczywistego silosu. Na podstawie przeprowadzonych analiz numerycznych zaproponowano uproszczone modele numeryczne silosów w formie segmentu silosu lub pojedynczego słupa na podłożu sprężystym. Zaproponowane modele pozwalają na uzyskanie wyników analiz stateczności i nośności granicznej zgodnych z wynikami modelu powłokowego całego silosu. Zaproponowano również modyfikacje procedur normowych pozwalających na dokładniejsze w porównaniu z aktualnymi procedurami normowymi, określenie nośności wyboczeniowej analizowanych silosów metodami analitycznymi. W pracy omówiono także wyniki doświadczeń dotyczących wyboczenia słupa silosowego w skali laboratoryjnej oraz w skali rzeczywistej.
Urban flash flood hazard identification and assessment applying geospatial techniques and hydrodynamic modeling; Erbil city case study, Kurdistan Region of Iraq
- Andam Mustafa
2023 Pełny tekst
This dissertation aims to investigate the factors behind flash flooding in Erbil's central district, located in the Kurdistan Region of Iraq, and develop a methodology for assessing flood hazards in the city, despite limited data accessibility. In this thesis, each factor was investigated, including analyzing extreme precipitation events in the last two decades, including their spatial and temporal distribution of rainfall, intensity, and exceedance probability, and examines the impact of changes in Land Use and Land Cover (LULC) on the hydrological response of the Erbil basin. The hydrodynamic model's input data were generated using GIS-based modeling interface. HEC-RAS 2-D software package's suitability was ensured by evaluating two building representation techniques and two mathematical models (Diffusion-Wave Equations (DWE) and Shallow-Water Equations (SWE)) using the Toce River urban flood experimental model. The study utilized a two-dimensional hydrodynamic model (HEC-RAS 2-D) to assess the susceptibility, vulnerability, and socioeconomic impact of flooding in the study area. Using the model, flood hazard maps were created to show the extent of potential flooding in the study area during various rainfall events and return periods. Ultimately, the study concludes that without essential engineering measures, there is an increased probability of flooding in the center of Erbil.
Urban scene semantic segmentation using the U-Net model
- Marcin Ciecholewski
2023
Vision-based semantic segmentation of complex urban street scenes is a very important function during autonomous driving (AD), which will become an important technology in industrialized countries in the near future. Today, advanced driver assistance systems (ADAS) improve traffic safety thanks to the application of solutions that enable detecting objects, recognising road signs, segmenting the road, etc. The basis for these functionalities is the adoption of various classifiers. This publication presents solutions utilising convolutional neural networks, such as MobileNet and ResNet50, which were used as encoders in the U-Net model to semantically segment images of complex urban scenes taken from the publicly available Cityscapes dataset. Some modifications of the encoder/decoder architecture of the U-Net model were also proposed and the result was named the MU-Net. During tests carried out on 500 images, the MU-Net model produced slightly better segmentation results than the universal MobileNet and ResNet networks, as measured by the Jaccard index, which amounted to 88.85\%. The experiments showed that the MobileNet network had the best ratio of accuracy to the number of parameters used and at the same time was the least sensitive to unusual phenomena occurring in images.
Usability study of various biometric techniques in bank branches
- Arkadiusz Harasimiuk
- Andrzej Czyżewski
2023 Pełny tekst
The purpose of the presented research was to evaluate the performance of the prepared biometric algorithms and obtain information on the opinions and preferences of their users in bank branches. The study aimed to determine users' attitudes towards particular modalities and preferences on how to use biometrics after the bank customers had practical experience with the operation of the prototype solutions. The research results allow a better understanding of users' needs and expectations, which can help to increase the knowledge of the need for biometric solutions in banking. Making appropriate changes based on the research results can help improve user satisfaction. The experiments included the collection of biometric sample collection and verification processes conducted by bank customers assisted by employees. The study collected biometric samples and questionnaires from 365 people, and 179 complete verifications were conducted. About 95% of those surveyed expressed acceptance of the implemented solutions for handling transactions using biometrics. The HandVein-FaceImage (hand vein scanning and face recognition) modality pair was rated highest in 70% of cases in the studied population.
Use of the AHP Method for Preference Determination in Yacht Design
- Jan Sierzputowski
- Artur Karczewski
- Przemysław Krata
2023 Pełny tekst Polish Maritime Research
A sailing yacht is a human-centred product, the design of which revolves primarily around the wants and desires of the future owner. In most cases, these preferences are not measurable, such as a personal aesthetic feeling, or a need for comfort, speed, safety etc. The aims of this paper are to demonstrate that these preferences can be classified and represented numerically, and to show that they are correlated with the type of yacht owned. As a case study, the owner’s preferences for deck equipment are considered. These are determined by pairwise comparisons of the importance rankings for features previously defined by yacht owners, following the analytic hierarchy process (AHP) method. As a result, a quantitative representation of these preferences is established, and they are shown to be correlated with the type of yacht. The findings of the current study show that the yacht owners’ preferences can be represented numerically, leading to a utilitarian conclusion that concerns the support and even some degree of automation of the design process.
User Authentication by Eye Movement Features Employing SVM and XGBoost Classifiers
- Piotr Odya
- Franciszek Górski
- Andrzej Czyżewski
2023 Pełny tekst IEEE Access
Devices capable of tracking the user’s gaze have become significantly more affordable over the past few years, thus broadening their application, including in-home and office computers and various customer service equipment. Although such devices have comparatively low operating frequencies and limited resolution, they are sufficient to supplement or replace classic input interfaces, such as the keyboard and mouse. The biometric application we researched verifies a user’s identity based on parameters acquired with a low-cost eye tracker. The use of the eye-tracking device in bank booths has many advantages, including the fact that eye trackers are contactless devices, which, especially in the light of the Covid pandemic, has increasing importance, in addition to providing a solution for confirming the liveness of the user. This paper describes an experiment in which 20 features extracted from eye movement data related mainly to saccades and fixations are used as a complementary biometric modality to authenticate clients at banking kiosks. Data were collected from 39 subjects while operating a banking system using engineered biometric kiosk prototypes. Authentication performance employing eye-movement tracking and parameterizing was compared for two classifiers: Support Vector Machine (SVM) and eXtreme Gradient Boosting (XGBoost). The results showed that the XGBoost-based classifier outperformed the SVM-based one regarding equal error rates (6.76% to 8% vs. 16.21 to 18.78%). Similar differences were obtained for true acceptance rates at different false acceptance rates (0.1 and 0.01), where the SVM-based classifier achieved a maximum of 81.08% and the XGBoost-based achieved 98.65%. Finally, prospects for the broader application of eye movement tracking as a biometric modality are discussed.
USING ARTIFICIAL NEURAL NETWORKS FOR PREDICTING SHIP FUEL CONSUMPTION
- Giao Van Nguyen
- Rajamohan Sakthivel
- Krzysztof Rudzki
- Janusz Kozak
- Sharma Prabhakar
- Nguyen Dang Khoa Pham
- Phuoc Quy Phong Nguyen
- Nguyen Xuan Phuong
2023 Pełny tekst Polish Maritime Research
In marine vessel operations, fuel costs are major operating costs which affect the overall profitability of the maritime transport industry. The effective enhancement of using ship fuel will increase ship operation efficiency. Since ship fuel consumption depends on different factors, such as weather, cruising condition, cargo load, and engine condition, it is difficult to assess the fuel consumption pattern for various types of ships. Most traditional statistical methods do not consider these factors when predicting marine vessel fuel consumption. With technological development, different statistical models have been developed for estimating fuel consumption patterns based on ship data. Artificial Neural Networks (ANN) are some of the most effective artificial methods for modelling and validating marine vessel fuel consumption. The application of ANN in maritime transport improves the accuracy of the regression models developed for analysing interactive relationships between various factors. The present review sheds light on consolidating the works carried out in predicting ship fuel consumption using ANN, with an emphasis on topics such as ANN structure, application and prediction algorithms. Future research directions are also proposed and the present review can be a benchmark for mathematical modelling of ship fuel consumption using ANN
Using Continuous Integration Techniques in Open Source Projects – An Exploratory Study
- Michał Wróbel
- Jarosław Szymukowicz
- Paweł Weichbroth
2023 Pełny tekst IEEE Access
For a growing number of software projects, continuous integration (CI) techniques are becoming an essential part of the process. However, the maturity of their adoption in open source projects varies. In this paper, we present an empirical study on GitHub repositories to explore the use of continuous integration techniques in open source projects. Following the Goal-Question-Metric (GQM) approach, 3 research questions and 7 metrics were defined for such a goal. We mined 10 repositories of open source projects with 101,149 pull requests, 399,671 commits from 20,432 developers. This was followed by exploratory data analysis for each metric. In summary, our results indicate that (RQ1) most failed CI builds required a small change in the pull request to fix the code; (RQ2) CI builds of smaller pull requests are more likely to succeed than larger ones; (RQ3) there was no correlation found between developer experience in committing to the project and the success rate of CI builds. Most of the projects studied have not yet developed a mature approach to using continuous integration techniques. In these cases, developers do not thoroughly test code before submitting pull requests. Furthermore, the results of the study confirmed that developers tend to submit pull requests with small amounts of new or modified code.
Using creative approaches for discovering biomorphic forms for appropriate human habitation in natural environments. Case study of Kashubian Lake District
- Agnieszka Kurkowska
2023 Pełny tekst Cogent Arts & Humanities
The research process consisted of studies of natural and cultural conditions of the Kashubian Lake District This is an area of exceptional natural conditions. For centuries, it has seen human habitation with respect to landscape values. Given its extensive forest cover and the lack of heterogeneity of natural conditions, the area has become an interesting inspiration for the author’s original project. The project is aimed at searching for appropriate forms of human habitation. Creative activities, owing to the creation of a large number of drawings and paintings that feature diagrams of initial forms, followed by sculptures inspired by them, aim at identifying biomorphic solutions, spatial forms that match the site of the study. Methodically, in order to achieve the goal, an original research and artistic concept was designed whose non-obvious results make it possible to treat the given task as a divergence problem. The essential artistic part comprised creating small-scale sculptures of clay, which were later to be placed in appropriately selected places. The photographic documentation offered the possibility to work at the analytical and comparative level, with an account of the obtained effects and their potential. Ultimately, the author seeks an answer to how to design houses in this region. The results come in the form of a complete multi-element set of biomorphic forms; directly or indirectly transformable into architectural objects, they provide model hints to classifiable sets of features conducive to arriving at a match with a place. The creative method proves appropriate for architectural design, as it constitutes its initial stage in the case of inhabiting natural environments. It yields favourable forecasts for creating a catalogue of design guidelines for residential buildings located in the area of the Kashubian Lake District.
Using Isolation Forest and Alternative Data Products to Overcome Ground Truth Data Scarcity for Improved Deep Learning-based Agricultural Land Use Classification Models
- Agustín Pereira García
- Lukasz Porwol
- Adegboyega Ojo
2023
High-quality labelled datasets represent a cornerstone in the development of deep learning models for land use classification. The high cost of data collection, the inherent errors introduced during data mapping efforts, the lack of local knowledge, and the spatial variability of the data hinder the development of accurate and spatially-transferable deep learning models in the context of agriculture. In this paper, we investigate the use of Isolation Forest (IF), an anomaly detection algorithm, to reduce noise in a large-scale, low-resolution alternative ground truth dataset used to train land use deep learning models. We use a modestsize, high-resolution and high-fidelity manually collected ground-truth dataset to calibrate Isolation Forest parameters and evaluate our approach, highlighting the relatively low cost of the methodology. Our data-centric methodology demonstrates the efficacy of deep learning methods coupled with IF to create mid-resolution land-use models and map products for agriculture using an alternative groundtruth dataset. Moreover, we compare our deep learning approach with a traditional algorithm used in remote sensing and evaluate the spatial transferability of the created models. Finally, we reflect upon the lessons learnt and future work