Repository of Faculty of Science
Not a member yet
1747 research outputs found
Sort by
Modeling and production of object by 3D printing
No full text3D ispis je proces izrade trodimenzionalnog objekta iz 3D modela ili drugog elektroničnog izvora podataka kroz aditivne procese u kojima se uzastopni slojevi materijala postavljaju pod kontrolom računala. U ovom radu opisane su različite tehnike 3D ispisa. Na primjeru LEGO kockice opisan je postupak izrade 3D modela u programu TinkerCAD. Nakon toga je prikazana izrada G koda korištenjem programa Creality Slicer i proces ispisa 3D modela na pisaču Ender-3 Pro. U zadnjem dijelu rada opisana su područja primjene 3D ispisne tehnologije.3D printing is the process of creating a three-dimensional object from a 3D model or other electronic data source through additive processes in which successive layers of material are placed with computer control. In this paper various 3D printing techniques are described. The procedure for creating a 3D model in the TinkerCAD program is described on the example of a LEGO brick. After that, the creation of G code using the Creality Slicer program and the process of printing a 3D model on the Ender-3 Pro printer were shown. In the last part of the paper, the areas of application of 3D printing technology are described
Difficulties in understanding basic programming concepts
No full textOvaj rad bavi se poteškoćama u osnovnim konceptima programiranja. Mnogi se suočavaju sa izazovom miskoncepcija, što često proizlazi iz nedovoljnog razumijevanja koncepata. Miskoncepcije mogu biti vezane uz različite aspekte programiranja, što rezultira nepotpunim ili neispravnim kodom. U ovome radu korišten je test, kojeg su ispunjavali studenti koji su pohađali kolegij OOP, kao i studenti koji se bave drugim oblicima programiranja. Istraživanje je provedeno u ožujku 2023. godine u učionicama Prirodoslovno-matematičkog fakulteta u Splitu, u kojem je sudjelovalo 223 studenata već spomenutog fakulteta. Analizom testa su identificirane miskoncepcije putem varijabli, grananja, petlji, nizova, funkcija i metoda. Ovi rezultati pružaju konkretnije uvide koji mogu pomoći pri unapređenju nastave programiranja, što je ključno za bolje razumijevanje osnovnih koncepata programiranja i pripremu studenata za izazove u stvarnom svijetu programiranja.This paper addresses the difficulties in understanding fundamental programming concepts. Many encounter the challenge of misconceptions, which often stem from insufficient understanding of various concepts. Misconceptions can be related to different aspects of programming, resulting in incomplete or faulty code. In this study, a test was used, filled out by students who attended the OOP course, as well as students engaged in other forms of programming. The research was conducted in March 2023 in classrooms at the Faculty of Science, University of Split, with the participation of 223 students from the aforementioned faculty. Various misconceptions were identified through the analysis of the test, including variables, branching, loops, arrays, functions, and methods. These results provide more specific insights that can aid in improving programming education, which is crucial for better understanding of basic programming concepts and preparing students for challenges in the real-world programming
K-means algorithm in bioinformatics
No full textU diplomskom radu predstavljen je k-means algoritam te, na primjeru analize ekspresije gena, njegova primjena u bioinformatici. U tu svrhu prvo smo predstavili potrebne pojmove iz biologije. Na kraju smo predstavili Lloydov algoritam kao jedan od najpoznatijih k-means algoritama. Također, u radu je dana i jedna implementacija Lloydovog algoritma za koju je korišten programski jezik Python.In this master’s thesis k-means algorithm is introduced, along its application in bioinformatics with an example of gene expression analysis. To achieve this, the necessary terms and concepts from biology are introduced first. At the end, Lloyd’s algorithm is presented as one of the most popular k-means algorithm. Additionally, one of the possible implementations of Lloyd’s algorithm is given in this thesis, for which programming language Python was used
Uloga specifičnih strukturnih motiva u načinu djelovanja antimikrobnog peptide TSO8 iz Taenia solium
No full textCombating antimicrobial resistance remains a major challenge in the recent decades, driving the search for innovative treatments. Antimicrobial peptides have emerged as a promising alternative, given their important role in defence mechanism of various organisms. This thesis focuses on TSO8 peptide identified in the helminth parasite Taenia solium, which is potent against both Gram-negative and Gram-positive bacterial strains. To assess the atomic details of its mode of action, TSO8 and four of its fragments, along with its linearized form, were investigated through all-atom molecular dynamics simulations in various environments. Results of the simulations reveal the significance of the initial GW amino acid sequence in stabilizing the α-helical structure. In interaction with the bacterial membrane model, two distinct contributions of peptide regions are observed: the N-terminal α-helix is crucial for the initial electrostatic contact, while the function of the unstructured C-terminal is most likely emphasized in other aspects of the peptide’s activity. Additionally, peptides show a high affinity of clustering in water, while exhibiting no significant clustering on the membrane surface. For further investigation the insight of MD simulations will be coupled with the biological and biophysical characterization on TSO8 and its fragments.Borba protiv antimikrobne rezistencije jedan je od glavnih izazova posljednjih desetljeća, potičući potragu za inovativnim liječenjima. Antimikrobni peptidi pojavili su se kao obećavajuća alternativa, obzirom na njihovu važnu ulogu u obrambenom mehanizmu raznih organizama. U ovom radu istražuje se peptid TSO8 identificiran u helmintskom parazitu Taenia solium, koji ima značajnu antimikrobnu aktivnost protiv Gram pozitivnih i Gram negativnih bakterijskih sojeva. Detalji mehanizma su istraženi simulacijama molekularne dinamike TSO8 peptida, četiri njegova fragmenta i njegova linearnog oblika, u različitim okruženjima. Rezultati simulacija otkrivaju važnost početnog GW slijeda aminokiselina u stabilizaciji α-helikalne strukture. U interakciji peptida s bakterijskom membranom ističu se dva različita doprinosa: N-terminalna α-uzvojnica ključna je za početni elektrostatski kontakt, dok je funkcija neuređenog C-terminala vjerojatno istaknuta u drugim aspektima aktivnosti peptida. Također, peptidi pokazuju visoku sklonost formiranju klastera u vodi, dok ne pokazuju značajno klasteriranje na površini membrane. U nastavku istraživanja rezultati simulacija će se upotpuniti eksperimentima za biološku i biofizičku karakterizaciju TSO8 peptida i njegovih fragmenata
Teaching search algorithms to primary school students using the CS Unplugged method
No full textDiplomski rad bavi se istraživanjem primjene metode "CS Unplugged" u podučavanju algoritama pretraživanja učenika razredne nastave, bez korištenja računala. Cilj rada bio je ispitati učinkovitost ovog pristupa u odnosu na tradicionalne metode podučavanja, s fokusom na usvajanje računalnih koncepata i motivaciju učenika. Istraživanje je provedeno na uzorku učenika trećeg razreda, koji su podijeljeni u eksperimentalnu skupinu (koja je učila algoritme pretraživanja metodom bez računala) i kontrolnu skupinu (koja je koristila tradicionalni pristup). Rezultati su pokazali da učenici iz eksperimentalne skupine postižu bolje rezultate na post-testovima te imaju višu razinu motivacije u odnosu na kontrolnu skupinu. Iako je uočeno blago poboljšanje retencije znanja na ponovljenim post-testovima, razlika nije statistički značajna. Zaključak rada upućuje na to da metoda "CS Unplugged" može biti koristan alat u podučavanju apstraktnih računalnih pojmova, potičući pritom kritičko razmišljanje i angažman učenika.The thesis explores the application of the "CS Unplugged" method in teaching search algorithms to elementary school students without the use of computers. The aim of the research was to examine the effectiveness of this approach compared to traditional teaching methods, with a focus on students' understanding of computational concepts and their motivation. The study was conducted on a sample of third-grade students, divided into an experimental group (which learned search algorithms using the unplugged method) and a control group (which followed a traditional teaching approach). The results showed that students in the experimental group achieved better results on post-tests and exhibited higher motivation levels compared to the control group. Although a slight improvement in knowledge retention was observed on the repeated post-tests, the difference was not statistically significant. The thesis concludes that the "CS Unplugged" method can be a useful tool for teaching abstract computational concepts, while also encouraging critical thinking and student engagement
Formacije strukturne boje od lipidnih liotropnih tekućih kristala
No full textNature employs structural color for purposes such as camouflage, attraction, and communication [1]. Despite progress in understanding the photonic structures responsible for structural color, the developmental pathways leading to these formations are not well understood [2]. How does a system transition from vesicles to a network-like structure, and what role does sugar play in this process? Is it simply a glue, or does it actively interact with vesicles to influence their organization and function? This project explores these questions using lipidic lyotropic liquid crystals to generate such structures. We use a dissipative particle dynamics (DPD) simulation model to predict and design biological building blocks for structural colors. DPD’s coarse-graining technique enables accurate simulations over larger particle numbers and longer timescales. Our model was tested for spatial and time convergence, ensuring reliability. Visualized trajectories reveal distinct phases with clarity, providing insight into structural evolution. The model also generated phase diagrams for various lipids, which may play a key role in structural coloration in many arthropods. This work advances our understanding of photonic structures and opens avenues for bio-inspired materials engineering.Priroda koristi strukturalne boje za kamuflažu, privlačenje i komunikaciju [1]. Unatoč napretku u razumijevanju fotoničkih struktura odgovornih za strukturalne boje, razvojni putevi koji vode do ovih formacija nisu dobro shvaćeni [2]. Kako sustav prelazi iz vezikula u mrežastu strukturu i koja je uloga šećera u tom procesu? Djeluje li samo kao ljepilo ili aktivno djeluje na vezikule, utječući na njihovu organizaciju i funkciju? Ovaj projekt istražuje ta pitanja koristeći lipidne liotropne tekuće kristale za stvaranje takvih struktura. Koristimo model simulacije diskipativne dinamičke čestice (DPD) za predviđanje i dizajniranje bioloških građevnih blokova za strukturalne boje. Tehnika grubog zrna DPD-a omogućuje točne simulacije na većem broju čestica i dužim vremenskim razmacima. Naš model testiran je za prostornu i vremensku konvergenciju, osiguravajući pouzdanost. Vizualizirane putanje jasno otkrivaju različite faze, pružajući uvid u strukturnu evoluciju. Model je također generirao fazne dijagrame za različite lipide, koji mogu igrati ključnu ulogu u strukturalnom bojanju mnogih člankonožaca. Ovaj rad unapređuje naše razumijevanje fotoničkih struktura i otvara put za bio-inspirirano inženjerstvo novih materijala
Modeling of interatomic interactions
No full textU radu je opisano modeliranje neutralnih bozona malih energija na niskim temperaturama pomoću Schrödingerove jednadžbe. Predstavljeni su fizički modeli poznatih oblika potencijala za aproksimaciju stvarnih interakcija među česticama. Interakcije čestica opisane su pomoću mjerljivih veličina iz teorije raspršenja: duljine raspršenja i efektivnog dosega. Podaci o interakcijama niskoenergijskih hiperfinih stanja izotopa pod različitim jakostima magnetskog polja preuzeti su iz ranijih istraživanja. Interakcije su aproksimirane primjenom triju različitih modela potencijala: pravokutna jama, pravokutna jama s pravokutnom barijerom i Lennard-Jonesov potencijal (10-6). U radu su predstavljene numeričke metode za rješavanje složenih diferencijalnih jednadžbi te algoritmi za efikasno pronalaženje parametara potencijala. Konačni rezultat istraživanja je tablica parametara potencijala za fizički model koji s visokom preciznošću određuje mjerljive parametre raspršenja. Predložene metode i algoritmi mogu poslužiti kao smjernice za buduća istraživanja u određivanju fizičkih modela ili kao osnova za primjenu izmjerenih podataka na složenije sustave s više čestica.The paper describes the modeling of neutral bosons at low energies and low temperatures using the Schrödinger equation. Physical models of known potential forms are presented to approximate real interactions between particles. The interactions are described using measurable quantities from scattering theory: scattering length and effective range. Data on the interactions of low-energy hyperfine states of isotopes under varying magnetic field strengths were taken from previous research. The interactions are approximated using three different potential models: square well, square well with square barrier, and Lennard-Jones potential (10-6). The paper presents numerical methods for solving complex differential equations and algorithms for efficiently determining potential parameters. The final outcome of the research is a table of potential parameters for a physical model that precisely determines measurable scattering parameters. The proposed methods and algorithms can serve as guidelines for future research in determining physical models or as a basis for applying measured data to more complex multi-particle systems
Energy-size relation of quantum trimers in 2D
No full textOvaj rad istražuje osnovno energijsko stanje vezanog kvantnog sustava od tri čestice (trimera) u Lennard-Jones potencijalu. Primijenjene su dvije stohastičke metode: varijacijski Monte Carlo (VMC) i difuzijski Monte Carlo (DMC) algoritmi. VMC simulacija dala je stabilne rezultate i procijenila gornju granicu energije osnovnog stanja sustava. Utvrđeni su optimalni parametri probne korelacijske funkcije za nekoliko sustava, a detaljnije prezentirani za dva. Primjenom DMC algoritma sustav je postigao stanje niže energije, ali uz nešto veće fluktuacije srednjih vrijednosti u usporedbi s VMC simulacijom, što upućuje na moguće tehničke ili teorijske izazove. Istražena je ovisnost energije dobivene DMC algoritmom o imaginarnom vremenskom koraku te je procijenjeno stvarno osnovno stanje sustava. Mogući uzroci nepravilnosti primjećenih u DMC algoritmu uključuju neadekvatno odabrane parametre, poteškoće u kontroliranju fluktuacija energije i broja šetača, ili tehničke probleme u kodu. Koristeći rezultate iz obje simulacije, istražena je univerzalnost skaliranja energije i širine trimera u 2D. Zaključno, za precizniju procjenu osnovnog stanja trimera potrebno je povećati računalnu snagu, proširiti VMC simulaciju na veći broj parametara, te unaprijediti DMC algoritam dodavanjem čistih estimatora.This paper investigates the ground state energy of a bound quantum system consisting of three particles (trimer) in the Lennard-Jones potential. Two stochastic methods were applied: the variational Monte Carlo (VMC) and the diffusion Monte Carlo (DMC) algorithms. The VMC simulation yielded stable results and estimated the upper bound of the system’s ground state energy, with optimal parameters for the trial correlation function being determined for examples of two potentials. By applying the DMC algorithm, the system reached a lower energy state, but with somewhat larger fluctuations in average quantities compared to the VMC simulation, indicating potential technical or theoretical challenges. The dependence of the energy simulated by the DMC algorithm on the imaginary time step was studied, and the true ground state of the system was estimated. Possible causes include inadequately chosen parameters, difficulties in controlling energy fluctuations and walker numbers, or technical issues in the code. Using the results from VMC, the universality of the relation between energy and size of trimers in 2D has been investigated. In conclusion, to obtain a more accurate estimate of the trimer’s ground state, it is necessary to increase the computational power, expand the VMC simulation to cover a broader range of parameters, and improve the DMC algorithm by adding pure estimators
Anticancerous potential of transition metal complexes
No full textRaznolika svojstva kompleksnih spojeva prijelaznih metala koji uključuju koordinacijske brojeve, geometriju, redoks svojstva, višestruka oksidacijska stanja prijelaznih metala i sposobnost da vežu mnoštvo tipova liganada daje dobru podlogu za njihove medicinske primjene. Otkrićem cisplatina, anorganskog antikancerogenog lijeka, došlo je do snažnog razvoja bioanorganske kemije koja kontinuirano teži prema razvoju novih lijekova na bazi metala. Istraživanja kompleksa prijelaznih metala, njihovih svojstava i načela djelovanja, vodi do sljedećih generacija lijekova, ali i prevladavanja nekih nedostataka povezanih s postojećim lijekovima kao što su smanjenje nuspojava, rezistencije i povećanje njihovog raspona aktivnosti.A variety of properties transition metal complexes, including coordination numbers, geometry, redox behavior, multiple oxidation states of transition metals and their ability to bind a wide variety type of ligands offer a strong basis for their medicinal applications. Serendipitous discovery of cisplatin, inorganic anticancer drug, led to strong development of medicinal inorganic chemistry, which continuously working towards development of new metal-based drugs. Exploration of transition metal complexes, their properties and modes of action, lead to future generations of drugs, as well as overcoming some of disadvantages associated with present drugs, including the reduction of side-effects, resistance and increasing their range of activity
Gamification in learning programming through online quizzes
No full textCilj ovog diplomskog rada bio je istražiti koncept igrifikacije i njezinu primjenu u obrazovanju, kao i u drugim područjima. Rad obuhvaća pregled i usporedbu nekih od najpopularnijih alata za izradu online kvizova, ističući njihove karakteristike i funkcionalnosti. Ključni dio rada odnosi se na razvoj i izradu web aplikacije namijenjene učenju programiranja putem online kvizova, koju mogu koristiti i nastavnici i učenici. U radu su detaljno opisane korištene tehnologije te postupak implementacije aplikacije, s naglaskom na njezinu funkcionalnost i prilagodljivost obrazovnim potrebama.The objective of this thesis was to explore the concept of gamification and its application in education, as well as in other fields. The paper provides a review and comparison of some of the most popular tools for creating online quizzes, emphasizing their features and functionalities. The core part of the thesis focuses on the development and design of a web application intended for programming education through online quizzes, suitable for both teachers and students. The thesis thoroughly describes the technologies used and the process of implementing the application, with a focus on its functionality and adaptability to educational needs