Search the digital library catalog

Abstract: This paper proposes two hybrid approaches based on Autoencoders (AEs) for long-term temperature prediction. The first algorithm comprises an AE trained to learn temperature patterns, which is then linked to a second AE, used to detect possible anomalies and provide a final temperature prediction. The second proposed approach involves training an AE and then using the resulting latent space as input of a neural network, which will provide the final prediction output. Both approaches are tested in long-term air temperature prediction in European cities: seven European locations where major heat waves occurred have been considered. The longterm temperature prediction for the entire year of the heatwave events has been analysed. Results show that the proposed approaches can obtain accurate long-term (up to 4 weeks) temperature prediction, improving Persistence and Climatology in the benchmark models compared. In heatwave periods, where the persistence of the temperature is extremely high, our approach beat the persistence operator in three locations and works similarly in the rest of the cases, showing the potential of this AE-based method for long-term temperature prediction.
Keywords: autoencoder, temperature prediction, hybrid models, heatwave
Published in DKUM: 29.01.2025; Views: 0; Downloads: 2
Full text (1,82 MB)

Abstract: With the onset of the COVID-19 pandemic, social media became one of the most utilized sources of information relating to the disease. With the increased reliance on social media, the risk of excessive use and the development of social media addiction emerges. The aim of the present study was to explore the psychometric properties of the Slovenian version of the Bergen Social Media Addiction Scale, and to explore how psychological resilience affects social media addiction symptoms directly and indirectly through symptoms of depression, anxiety and mental distress. A large online cross-sectional study was conducted in March 2021 among Slovenian tertiary students (N = 4868). The results showed the high reliability, unidimensionality and criterion validity of the Slovenian Bergen Social Media Addiction Scale. The proposed structural model fit the data well and showed a significant direct positive effect of depression and stress on social media addiction. Moreover, the majority of the negative effects of psychological resilience on social media addiction (87.2%) were indirect, through depression and stress symptoms, whereas resilience had a significantly smaller impact on social media addiction by reducing anxiety symptoms. The overall prevalence of social media addiction symptoms was 4.6%, with females exhibiting higher proportions than men. Additionally, female social media users reported a complete absence of social media addiction symptoms less often compared to males. Future research should further explore the mechanisms behind social media addiction, in order to gain a better understanding of the apparently different risk levels for both genders.
Keywords: social media addiction, behavioral addictions, resilience, validation, Bergen Social Media Addiction Scale (BSMAS), COVID-19
Published in DKUM: 09.07.2024; Views: 128; Downloads: 22
Full text (1,35 MB)
This document has many files! More...

Abstract: Namen magistrske naloge je bil optimizirati formulacijo polimernih nosilcev kot sistemov za dostavo farmacevtskih učinkovin. Cilj je bil izboljšati topnost izbranih farmacevtskih učinkovin ter doseči podaljšano sproščanje v simuliranih telesnih tekočinah. Polimerni nosilci so obetavni na področju razvoja sistemov s formuliranimi zdravili, saj lahko z ustrezno prilagoditvijo njihovih fizikalno-kemijskih lastnosti dosežemo izboljšano formuliranje in sproščanje farmacevtskih učinkovin. V sklopu laboratorijskega dela smo sintetizirali plastovite polimerne nosilce, v katere smo uspešno formulirali dve različni farmacevtski učinkovini z različno polarnostjo. Izbrana polimera sta bila hitozan in alginat. S procesom superkritičnega penjenja smo izdelali tudi porozne polimerne nosilce, kjer smo kot začetni polimer uporabili polikaprolakton. Kot farmacevtski učinkovini smo izbrali ketoprofen in klindamicin hidroklorid. Izbrali smo ju glede na njuno topnost v telesnih tekočinah. Ketoprofen se uporablja za zdravljenje akutne bolečine in je slabo topen, medtem ko je klindamicin hidroklorid bolj topen v telesnih tekočinah in se uporablja kot antibiotik. Karakterizacijo formuliranih polimernih nosilcev z vsebnostjo farmacevtskih učinkovin smo izvedli z vrstično elektronsko mikroskopijo in tehniko oslabljenega totalnega odboja infrardeče spektroskopije s Fourierjevo transformacijo. Z in vitro testiranjem sproščanja smo prikazali spreminjanje vsebnosti formuliranih farmacevtskih učinkovin v polimernih nosilcih glede na čas sproščanja, pridobljene vzorce smo analizirali s tekočinsko kromatografijo visoke ločljivosti in z UV-Vis spektrofotometrijo. Rezultati so nedvoumno pokazali, da je formulacija farmacevtskih učinkovin v polimerne nosilce omogočila podaljšano sproščanje v telesnih tekočinah. Ugotovitve kažejo, da so porozni polimerni nosilci bolj primerni za sisteme dostave zdravil s podaljšanim sproščanjem v primerjavi s plastovitimi polimernimi nosilci.
Keywords: farmacevtske učinkovine, polimerni nosilci, plastoviti polimeri, superkritično penjenje, in vitro sproščanje
Published in DKUM: 03.04.2024; Views: 273; Downloads: 63
Full text (2,20 MB)

Abstract: V tej diplomski nalogi smo raziskali ogrodje Hadoop, ki s svojimi komponentami tvori celovito rešitev za hranjenje in analiziranje velikih podatkov. V diplomski nalogi najprej predstavimo ogrodje in njegove glavne oziroma najbolj uporabljene komponente, kot so HDFS, MapReduce in YARN. Sledi primer vzpostavitve ogrodja na Linux distribuciji Ubuntu, ter primeri uporabe, ki podajajo smernice za shranjevanje in analiziranje različnih vrst podatkov s Hadoop.
Keywords: Hadoop, vele podatki, porazdeljen datotečni sistem, shranjevanje in analiza podatkov
Published in DKUM: 05.10.2023; Views: 418; Downloads: 27
Full text (2,32 MB)

Abstract: Elektromagnetno polje obstaja povsod. V sklopu te diplomske naloge smo se osredotočili na polje, katerega izvor je 20 kV daljnovod. Polje v karakteristični ravnini smo prikazali z izvedbo meritve ter z izdelavo simulacije. Pridobljene vrednosti smo primerjali med seboj ter tudi s priporočili ICNIRP. Opisali smo probleme povezane z izvedbo tovrstnih meritev in simulacij. Meritve smo izvedli s senzorjem NFA 400 3D NF Analyser, simulacije pa s pomočjo brezplačnega programa FEMM 4.2.
Keywords: Analiza elektromagnetnega polja, meritev EMP, simulacija EMP, distribucijski daljnovod.
Published in DKUM: 09.07.2021; Views: 932; Downloads: 113
Full text (1,72 MB)