Yazar "Kurban, Hasan" seçeneğine göre listele

Listeleniyor 1 - 12 / 12
  • [ X ]
    Öğe
    A DFT study on stability and electronic structure of AlN nanotubes
    (Elsevier, 2021) Muz, Iskender; Kurban, Hasan; Kurban, Mustafa
    Structural, energetic, electronic, reactivity and stability properties of armchair (3,3), (4,4), (5,5), (6,6), (7,7), (8,8), (9,9) and (10,10) aluminum nitride nanotubes (AlNNTs) with different diameter have been probed using density functional theory (DFT) in terms of Moreover, the chemical reactivity characteristics of AlNNTs have performed via some of the quantum molecular descriptors. Our results also indicate that the increasing diameter of AlNNTs gives rise to notable changes in the electronic structure of the AlNNTs. Moreover, results for UV/vis spectra of AlNNTs indicate that the maximum wavelength absorption lie in the range 188-194 nm. The number Al-N bonds and segregation phenomena of Al and N atoms in the AlNNTs have been investigated to better understand the stability of AlNNTs. Besides, the energy gap and chemical hardness enhance with increase diameter of AlNNTs, thus resulting in a rise in the stability, while the AlNNTs with smaller can be considered as a candidate for the adsorption of gas molecules and drugs for nano-electronic applications.
  • [ X ]
    Öğe
    Atom classification with Machine Learning and correlations among physical properties of ZnO nanoparticle
    (Elsevier, 2021) Kurban, Hasan
    Machine Learning (ML) has been recently used to make sense of large volume of data as data-driven methods to identify correlations and then examine material properties in detail. Herein, we analyze the correlations between structural and electronic properties of ZnO nanoparticles (NPs) obtained from density-functional tight-binding method using Data Science techniques. More clearly, the Pearson correlation coefficients were first computed to perform the relationship among the physical properties of ZnO NPs. Second, we classified Zn and O atoms using optimized geometries of ZnO NPs at different temperatures using various of ML algorithms. Our results show that segregation phenomena and bonding of Zn-O and O-O two-body interactions have a stronger relationship with the orbital energies than that of Zn-Zn. We also observe that a specific type of ML algorithm, tree-based models, performs much better than other types. Additionally, Random Forest outperforms other algorithms and is able to learn ZnO NPs close to perfect.
  • [ X ]
    Öğe
    Building Machine Learning systems for multi-atoms structures: CH3NH3PbI3 perovskite nanoparticles
    (Elsevier, 2021) Kurban, Hasan; Kurban, Mustafa
    In this study, we built a variety of Machine Learning (ML) systems over 23 different sizes of CH3NH3PbI3 perovskite nanoparticles (NPs) to predict the atoms in the NPs from their geometric locations. Our findings show that a specific type of ML algorithms, tree-based models which are Random Forest (RF), Extreme Gradient Boosting (XGBoost), Decision Trees (DT), can perfectly learn CH3NH3PbI3 perovskite NPs. Surprisingly, some popular ML algorithms such as Naive Bayes (NB), Support Vector Machines (SVM), Partial Least Squares (PLS), Regularized Logistic Regression (LR), Neural Networks (NN), Stacked Auto-Encoder Deep Neural Network (DNN), K-Nearest Neighbor (KNN) fail to learn CH3NH3PbI3 perovskite NPs.
  • [ X ]
    Öğe
    ccImpute: an accurate and scalable consensus clustering based algorithm to impute dropout events in the single-cell RNA-seq data
    (Bmc, 2022) Malec, Marcin; Kurban, Hasan; Dalkilic, Mehmet
    Background: In recent years, the introduction of single-cell RNA sequencing (scRNA-seq) has enabled the analysis of a cell's transcriptome at an unprecedented granularity and processing speed. The experimental outcome of applying this technology is a M x N matrix containing aggregated mRNA expression counts of M genes and N cell samples. From this matrix, scientists can study how cell protein synthesis changes in response to various factors, for example, disease versus non-disease states in response to a treatment protocol. This technology's critical challenge is detecting and accurately recording lowly expressed genes. As a result, low expression levels tend to be missed and recorded as zero - an event known as dropout. This makes the lowly expressed genes indistinguishable from true zero expression and different than the low expression present in cells of the same type. This issue makes any subsequent downstream analysis difficult. Results: To address this problem, we propose an approach to measure cell similarity using consensus clustering and demonstrate an effective and efficient algorithm that takes advantage of this new similarity measure to impute the most probable dropout events in the scRNA-seq datasets. We demonstrate that our approach exceeds the performance of existing imputation approaches while introducing the least amount of new noise as measured by clustering performance characteristics on datasets with known cell identities. Conclusions: cclmpute is an effective algorithm to correct for dropout events and thus improve downstream analysis of scRNA-seq data. cclmpute is implemented in R and is available at https://github.com/khazum/ccImpute.
  • [ X ]
    Öğe
    DCEM: An R package for clustering big data via data-centric modification of Expectation Maximization
    (Elsevier, 2022) Sharma, Parichit; Kurban, Hasan; Dalkilic, Mehmet
    Clustering is intractable, so techniques exist to give a best approximation. Expectation Maximization (EM), initially used to impute missing data, is among the most popular. Parameters of a fixed number of probability distributions (PDF) together with the probability of a datum belonging to each PDF are iteratively computed. EM does not scale with data size, and this has hampered its current use. Using a data-centric approach, we insert hierarchical structures within the algorithm to separate high expressive data (HE) from low expressive data (LE): the former greatly affects the objective function at some iteration i, while LE does not. By alternating using either HE or HE+LE, we significantly reduce run-time for EM. We call this new, data-centric EM, EM*. We have designed and developed an R package called DCEM (Data Clustering with Expectation Maximization) to emphasize that data is driving the algorithm. DCEM is superior to EM as we vary size, dimensions, and separability, independent of the scientific domain. DCEM is modular and can be used as either a stand-alone program or a pluggable component. DCEM includes our implementation of the original EM as well. To the best of our knowledge, there is no open source software that specifically focuses on improving EM clustering without explicit parallelization, modified seeding, or data reduction. DCEM is freely accessible on CRAN (Comprehensive R Archive Network). (C) 2021 The Author(s). Published by Elsevier B.V.
  • [ X ]
    Öğe
    Density-functional tight-binding approach for the structural analysis and electronic structure of copper hydride metallic nanoparticles
    (Elsevier, 2019) Kurban, Hasan; Kurban, Mustafa; Dalkilic, Mehmet
    We perform a theoretical investigation using the density functional tight-binding (DFTB) approach for the structural analysis and electronic structure of copper hydride (CuH) metallic nanoparticles (NPs) of different size (from 0.7 to 1.6 nm). By increasing the size of CuH NPs, the number of bonds, segregation phenomena and radial distribution function (RDF) of binary Cu-Cu, Cu-H and H-H interactions are analyzed using new implementations in R code. The results reveal that the number of Cu-Cu bonds is more than that of Cu-H while the number of H-H bonds are the less. Thus, a large amount of H atoms prefers to connect to Cu atoms. The increase in the size of the NPs contributes to their stabilization because of the increase in the interaction of H-H bonding. The segregation of Cu and H atoms shows that Cu atoms tend to co-locate at the center, while H atoms tend to reside on the surface. From the density of state (DOS) analysis, CuH NPs shows a metallic character which is compatible with experimental data. HOMO and Fermi levels decrease from -3.555 to -3.443 eV and from -3.510 to -3.441 eV. Herein, an increase in the size contributes to the stabilization of CuH NP due to decrease in the HOMO energies.
  • [ X ]
    Öğe
    Effect of Mg content on electronic structure, optical and structural properties of amorphous ZnO nanoparticles: A DFTB study
    (Elsevier, 2021) Kurban, Hasan; Alaei, Sholeh; Kurban, Mustafa
    In this work, we perform a theoretical analysis of structural, electronic, and optical properties of pure and Mg-doped amorphous ZnO nanoparticles (a-ZnO NPs) using DFTB method. Our results show that Zn atoms are more preferential for Mg atoms than for O atoms because the number of Mg-Zn bonds is greater than that of Mg-O. The rise in the content of Mg in a-ZnO NPs leads to an increase of Mg-Zn and Mg-O interactions. Mg atoms prefer to locate near the center of a-ZnO NP, but Zn and O atoms nearly preserve their positions which is compatible with radial distribution function peaks. The orbital energies display a decrease in the energy gap from 3.592 to 3.546 eV while increasing Mg content. The LUMO level is also significantly shifted to higher energies. The results also reveal that the performance of pure a-ZnO NP can be enhanced with a subsequent increase in Mg content.
  • [ X ]
    Öğe
    Metin Madenciliği ile Tıbbi Tedavi Alanlarının Yakınlıklarının Ölçülmesi
    (2021) Kurban, Hasan
    Bazı hastalık belirtilerinin birçok tıbbi tedavi alanıyla ilgili olması, hastaların tedavi için randevu alırken zorlanmalarına sebep olmaktadır. Örneğin; karın ağrısı rahatsızlığı bulunan bir hastanın rahatsızlığı dahiliye, hariciye ya da intaniye bölümlerinden herhangi birisiyle ilgisi bulunabilmektedir. Bu çalışmada T.C. Sağlık Bakanlığına bağlı birçok kamu hastanesinin resmî internet sitesinde bulunan ve hastaların belirtilerine göre doğru tıbbi tedavi branşını seçmelerine yardımcı olmak amacıyla kullanılan 13 tıbbi alan ve 204 belirti, metin madenciliği ve veri bilimi teknikleriyle kapsamlı olarak incelenmiştir. Kamu hastanelerinin resmî internet sitelerde kullanılan metnin içeriği baz alınarak tıbbi tedavi alanları arasındaki, yakınlık/uzaklık hesaplanıp, kelime bazlı hastaları randevu alanını belirlerken en çok zorlayan kelimeler ve belirtiler tespit edilmiştir. Kullanılan kelimeler analiz edilirken edat ve bağlaç gibi anlamsız sözcükler göz ardı edilip, hastalık belirtileri üzerinde kelime bulutu (word cloud) oluşturulmuştur. Tıbbi alanların yakınlığını hesaplamak için öncelikle metin içeriği kullanılarak 13 alan için her bir belirtinin var olup olmadığını gösteren 13x186 boyutlu ikili veri (binary data, document matrix) oluşturulmuştur. Daha sonra, bu veri seti üzerinde tıbbi tedavi alanları belirtilere göre aglomeratif hiyerarşik kümeleme algoritmaları (single, complete, average, ward, mcquitty) kullanılarak kümelendirilip metin bazlı alanların birbiri ile yakınlığı tespit edilmiştir. Bu makalenin sonuçlar kısmında hastaları en çok zorlayan kelimeler ve tıbbi alanların metin bazlı yakınlıkları paylaşılmıştır. Elde edilen sonuçlar çerçevesinde kullanılan metnin sağlık uzmanları tarafından tekrar düzenlenmesinin, yanlış tıbbi branşlardan alınan randevu sayısının azaltılmasına katkısı olacaktır.
  • [ X ]
    Öğe
    Predicting atom types of anatase tio2 nanoparticles with machine learning
    (Trans Tech Publications Ltd, 2021) Kurban, Hasan; Kurban, Mustafa; Sharma, Parichit; Dalkilic, Mehmet M.
    Machine learning (ML) has recently made a major contribution to the fields of Material Science (MS). In this study, ML algorithms are used to learn atoms types over structural geometrical data of anatase TiO2 nanoparticles produced at different temperature levels with the densityfunctional tight-binding method (DFTB). Especially for this work, Random Forest (RF), Decision Trees (DT), K-Nearest Neighbor (KNN), Naïve Bayes (NB), which are among the most popular ML algorithms, were run to learn titanium (Ti) and oxygen (O) atoms. RF outperforms other algorithms, almost succeeding in learning this skewed data set close to perfect. The use of ML algorithms with datasets compatible with its mathematical design increases their learning performance. Therefore, we find it remarkable that a certain type of ML algorithm performs almost perfectly. Because it can help material scientists predict the behavior and structural and electronic properties of atoms at different temperatures. © 2021 Trans Tech Publications Ltd, Switzerland.
  • [ X ]
    Öğe
    Rare-class learning over Mg-doped ZnO nanoparticles
    (Elsevier, 2021) Kurban, Hasan; Kurban, Mustafa
    This interdisciplinary study is conducted to find answers to two important questions which researchers often face in Machine Learning (ML) and Material Science (MS) fields. In this work, we measure the performance of the most popular ML algorithms (more than a dozen) on rare-class learning problem and determine the best learning algorithm for atom type prediction over the Mg-doped ZnO nanoparticles data obtained from the densityfunctional tight-binding method. As a result, we observe that tree-based ML algorithms such as Extreme Gradient Boosting (XGB), Decision Trees (DT), Random Forest (RF), outperform other types of ML algorithms, e. g., cost-sensitive learning, prototype models, support vector machines, kernel methods, on both rare-class learning and atom type prediction.
  • [ X ]
    Öğe
    Regeneration of Lithium-ion battery impedance using a novel machine learning framework and minimal empirical data
    (Elsevier, 2022) Temiz, Selcuk; Kurban, Hasan; Erol, Salim; Dalkilic, Mehmet M.
    The use of Electrochemical Impedance Spectroscopy on rechargeable Lithium-ion battery characterization is an extensively recognized non-destructive procedure for both in-situ and ex-situ analyses. In an impedance measurement for a rechargeable battery, the oscillating current with an accompanying phase angle is the response for a potential perturbation. The proper evaluation of phase angle as a crucial impedance parameter, provides critical understanding of the status of the battery. Although fast and simple, impedance data is difficult to interpret. Using a novel data-centric Machine Learning framework (co-modeling) we demonstrate how to impute experimental data quickly, precisely, and inexpensively that agrees with wholly experimentally generated data. In particular, we predict the phase angle with 99.9% accuracy by training the minimal empirical impedance data. This approach demonstrates a potentially burgeoning field of Machine Learning experimental data imputation and the consequence of faster diagnostic and study of batteries.
  • [ X ]
    Öğe
    Tailoring the structural properties and electronic structure of anatase, brookite and rutile phase TiO2 nanoparticles: DFTB calculations
    (Elsevier, 2020) Kurban, Hasan; Dalkilic, Mehmet; Temiz, Selcuk; Kurban, Mustafa
    In this study, we perform a theoretical investigation using the density functional tight-binding (DFTB) approach for the structural analysis and electronic structure of anatase, brookite and rutile phase TiO2 nanoparticles (NPs). Our results show that the number of Ti-O bonds is greater than that of O-O, while the number of Ti-Ti bonds is fewer. Thus, large amounts of O atoms prefer to connect to Ti atoms. The increase in the temperature of the NPs contributes to an increase in the interaction of Ti-O bonding, but a decrease in the O-O bonding. The segregation of Ti and O atoms shows that Ti atoms tend to co-locate at the center, while O atoms tend to reside on the surface. Increasing temperature causes a decrease of the bandgap from 3.59 to 2.62 eV for the brookite phase, which is much more energetically favorable compared to the bulk, while it could increase the bandgap from 3.15 to 3.61 eV for anatase phase. For three-phase TiO2 NPs, LUMO and Fermi levels decrease. The HOMO level of anatase phase NP decreases, but it increases for brookite and rutile phase TiO2 nanoparticles. An increase in the temperature contributes to the stabilization of anatase phase TiO2 NP due to a decrease in the HOMO energies.