-
A Comprehensive Study of Czernik 41 and NGC 1342 Using CCD UBV and Gaia DR3 Data
Authors:
B. Tanik Ozturk,
S. Bilir,
T. Yontan,
O. Plevne,
T. Ak,
S. Ak,
R. Canbay,
T. Banks
Abstract:
In this study, the structural, astrophysical, kinematic, and Galactic orbital parameters of the open clusters Czernik 41 and NGC 1342, as well as their dynamical evolution, are investigated using CCD UBV photometry and Gaia data. By applying the UPMASK algorithm to Gaia astrometric data for the estimation of cluster membership probabilities, we have determined that 382 stars in Czernik 41 and 111…
▽ More
In this study, the structural, astrophysical, kinematic, and Galactic orbital parameters of the open clusters Czernik 41 and NGC 1342, as well as their dynamical evolution, are investigated using CCD UBV photometry and Gaia data. By applying the UPMASK algorithm to Gaia astrometric data for the estimation of cluster membership probabilities, we have determined that 382 stars in Czernik 41 and 111 stars in NGC 1342 exhibit the highest statistical likelihood of being cluster members. Fundamental parameters (including reddening, metallicity, distance, and age) were derived using both classical methods, where parameters are determined separately, and Markov Chain Monte Carlo (MCMC) methods, where parameters are estimated simultaneously. The results obtained from both approaches are in agreement, confirming the reliability of the derived parameters and demonstrating their robustness against potential degeneracies. The distances to Czernik 41 and NGC 1342 were determined as 2485$\pm$151 pc and 645$\pm$42 pc, respectively, while their ages were estimated to be 69$\pm$10 Myr and 1000$\pm$50 Myr. The metallicity values ([Fe/H]) were found to be 0.07$\pm$0.09 dex for Czernik 41 and -0.14$\pm$0.07 dex for NGC 1342. The stellar mass functions for both clusters were derived, yielding slopes of $Γ$=1.67$\pm$0.23 for Czernik 41 and $Γ$ =1.56$\pm$0.41 for NGC 1342. Kinematic orbit analysis indicates that Czernik 41 originated within the Solar circle, whereas NGC 1342 formed outside it.
△ Less
Submitted 7 July, 2025;
originally announced July 2025.
-
Efficient Tree Generation for Globally Optimal Decisions under Probabilistic Outcomes
Authors:
Berk Ozturk,
She'ifa Punla-Green,
Les Servi
Abstract:
Many real-world problems require making sequences of decisions where the outcomes of each decision are probabilistic and uncertain, and the availability of different actions is constrained by the outcomes of previous actions. There is a need to generate policies that are adaptive to uncertainty, globally optimal, and yet scalable as the state space grows. In this paper, we propose the generation o…
▽ More
Many real-world problems require making sequences of decisions where the outcomes of each decision are probabilistic and uncertain, and the availability of different actions is constrained by the outcomes of previous actions. There is a need to generate policies that are adaptive to uncertainty, globally optimal, and yet scalable as the state space grows. In this paper, we propose the generation of optimal decision trees, which dictate which actions should be implemented in different outcome scenarios, while maximizing the expected reward of the strategy. Using a combination of dynamic programming and mixed-integer linear optimization, the proposed methods scale to problems with large but finite state spaces, using problem-specific information to prune away large subsets of the state space that do not yield progress towards rewards. We demonstrate that the presented approach is able to find the globally optimal decision tree in linear time with respect to the number states explored.
△ Less
Submitted 24 April, 2025;
originally announced April 2025.
-
Bridging vision language model (VLM) evaluation gaps with a framework for scalable and cost-effective benchmark generation
Authors:
Tim Rädsch,
Leon Mayer,
Simon Pavicic,
A. Emre Kavur,
Marcel Knopp,
Barış Öztürk,
Klaus Maier-Hein,
Paul F. Jaeger,
Fabian Isensee,
Annika Reinke,
Lena Maier-Hein
Abstract:
Reliable evaluation of AI models is critical for scientific progress and practical application. While existing VLM benchmarks provide general insights into model capabilities, their heterogeneous designs and limited focus on a few imaging domains pose significant challenges for both cross-domain performance comparison and targeted domain-specific evaluation. To address this, we propose three key c…
▽ More
Reliable evaluation of AI models is critical for scientific progress and practical application. While existing VLM benchmarks provide general insights into model capabilities, their heterogeneous designs and limited focus on a few imaging domains pose significant challenges for both cross-domain performance comparison and targeted domain-specific evaluation. To address this, we propose three key contributions: (1) a framework for the resource-efficient creation of domain-specific VLM benchmarks enabled by task augmentation for creating multiple diverse tasks from a single existing task, (2) the release of new VLM benchmarks for seven domains, created according to the same homogeneous protocol and including 162,946 thoroughly human-validated answers, and (3) an extensive benchmarking of 22 state-of-the-art VLMs on a total of 37,171 tasks, revealing performance variances across domains and tasks, thereby supporting the need for tailored VLM benchmarks. Adoption of our methodology will pave the way for the resource-efficient domain-specific selection of models and guide future research efforts toward addressing core open questions.
△ Less
Submitted 21 February, 2025;
originally announced February 2025.
-
Experimental Characterization of Hydrodynamic Gating-Based Molecular Communication Transmitter
Authors:
Eren Akyol,
Ahmet Baha Ozturk,
Iman Mokari Bolhassan,
Murat Kuscu
Abstract:
Molecular communication (MC) is a bio-inspired method of transmitting information using biochemical signals, promising for novel medical, agricultural, and environmental applications at the intersection of bio-, nano-, and communication technologies. Developing reliable MC systems for high-rate information transfer remains challenging due to the complex and dynamic nature of application environmen…
▽ More
Molecular communication (MC) is a bio-inspired method of transmitting information using biochemical signals, promising for novel medical, agricultural, and environmental applications at the intersection of bio-, nano-, and communication technologies. Developing reliable MC systems for high-rate information transfer remains challenging due to the complex and dynamic nature of application environments and the physical and resource limitations of micro/nanoscale transmitters and receivers. Microfluidics can help overcome many such practical challenges by enabling testbeds that can replicate the application media with precise control over flow conditions. However, existing microfluidic MC testbeds face significant limitations in chemical signal generation with programmable signal waveforms, e.g., in terms of pulse width. To tackle this, we previously proposed a practical microfluidic MC transmitter architecture based on the hydrodynamic gating technique, a prevalent chemical waveform generation method. This paper reports the experimental validation and characterization of this method, examining its precision in terms of spatiotemporal control on the generated molecular concentration pulses. We detail the fabrication of the transmitter, its working mechanism and discuss its potential limitations based on empirical data. We show that the microfluidic transmitter is capable of providing precise, programmable, and reproducible molecular concentration pulses, which would facilitate the experimental research in MC.
△ Less
Submitted 19 August, 2024;
originally announced August 2024.
-
Fair by design: A sociotechnical approach to justifying the fairness of AI-enabled systems across the lifecycle
Authors:
Marten H. L. Kaas,
Christopher Burr,
Zoe Porter,
Berk Ozturk,
Philippa Ryan,
Michael Katell,
Nuala Polo,
Kalle Westerling,
Ibrahim Habli
Abstract:
Fairness is one of the most commonly identified ethical principles in existing AI guidelines, and the development of fair AI-enabled systems is required by new and emerging AI regulation. But most approaches to addressing the fairness of AI-enabled systems are limited in scope in two significant ways: their substantive content focuses on statistical measures of fairness, and they do not emphasize…
▽ More
Fairness is one of the most commonly identified ethical principles in existing AI guidelines, and the development of fair AI-enabled systems is required by new and emerging AI regulation. But most approaches to addressing the fairness of AI-enabled systems are limited in scope in two significant ways: their substantive content focuses on statistical measures of fairness, and they do not emphasize the need to identify and address fairness considerations across the whole AI lifecycle. Our contribution is to present an assurance framework and tool that can enable a practical and transparent method for widening the scope of fairness considerations across the AI lifecycle and move the discussion beyond mere statistical notions of fairness to consider a richer analysis in a practical and context-dependent manner. To illustrate this approach, we first describe and then apply the framework of Trustworthy and Ethical Assurance (TEA) to an AI-enabled clinical diagnostic support system (CDSS) whose purpose is to help clinicians predict the risk of developing hypertension in patients with Type 2 diabetes, a context in which several fairness considerations arise (e.g., discrimination against patient subgroups). This is supplemented by an open-source tool and a fairness considerations map to help facilitate reasoning about the fairness of AI-enabled systems in a participatory way. In short, by using a shared framework for identifying, documenting and justifying fairness considerations, and then using this deliberative exercise to structure an assurance case, research on AI fairness becomes reusable and generalizable for others in the ethical AI community and for sharing best practices for achieving fairness and equity in digital health and healthcare in particular.
△ Less
Submitted 13 June, 2024;
originally announced June 2024.
-
Enhancements to the BOUN Treebank Reflecting the Agglutinative Nature of Turkish
Authors:
Büşra Marşan,
Salih Furkan Akkurt,
Muhammet Şen,
Merve Gürbüz,
Onur Güngör,
Şaziye Betül Özateş,
Suzan Üsküdarlı,
Arzucan Özgür,
Tunga Güngör,
Balkız Öztürk
Abstract:
In this study, we aim to offer linguistically motivated solutions to resolve the issues of the lack of representation of null morphemes, highly productive derivational processes, and syncretic morphemes of Turkish in the BOUN Treebank without diverging from the Universal Dependencies framework.
In order to tackle these issues, new annotation conventions were introduced by splitting certain lemma…
▽ More
In this study, we aim to offer linguistically motivated solutions to resolve the issues of the lack of representation of null morphemes, highly productive derivational processes, and syncretic morphemes of Turkish in the BOUN Treebank without diverging from the Universal Dependencies framework.
In order to tackle these issues, new annotation conventions were introduced by splitting certain lemmas and employing the MISC (miscellaneous) tab in the UD framework to denote derivation. Representational capabilities of the re-annotated treebank were tested on a LSTM-based dependency parser and an updated version of the BoAT Tool is introduced.
△ Less
Submitted 24 July, 2022;
originally announced July 2022.
-
A study of the NGC 1193 and NGC 1798 open clusters using CCD UBV photometric and Gaia EDR3 data
Authors:
T. Yontan,
H. Cakmak,
S. Bilir,
T. Banks,
M. Raul,
R. Canbay,
S. Koc,
S. Tasdemir,
H. Ercay,
B. Tanik Ozturk,
D. C. Dursun
Abstract:
We present photometric, astrometric, and kinematic studies of the old open star clusters NGC 1193 and NGC 1798. Both of the clusters are investigated by combining data sets from Gaia EDR3 and CCD UBV observational data. Analysis of the radial distribution of stars through the cluster regions indicates that the cluster limit radii are $r_{\rm lim}=8'$ for both of the clusters. We determine the memb…
▽ More
We present photometric, astrometric, and kinematic studies of the old open star clusters NGC 1193 and NGC 1798. Both of the clusters are investigated by combining data sets from Gaia EDR3 and CCD UBV observational data. Analysis of the radial distribution of stars through the cluster regions indicates that the cluster limit radii are $r_{\rm lim}=8'$ for both of the clusters. We determine the membership probabilities of stars considering Gaia EDR3 proper motion and trigonometric parallax data, resulting in 361 stars in NGC 1193 and 428 in NGC 1798 being identified as most likely cluster members, having membership probabilities greater than P>0.5. Mean proper motion components are estimated as ($μ_α\cos δ$, $μ_δ) = (-0.207(0.009), -0.431(0.008)$) for NGC 1193 and ($μ_α\cos δ$, $μ_δ)=(0.793(0.006), -0.373(0.005)$) mas/yr for NGC 1798. E(B-V) color excesses were derived for NGC 1193 as $0.150(0.037)$ and for NGC 1798 as 0.505(0.100) mag through the use of two-color diagrams. Photometric metallicities are also determined from two-color diagrams with the results of [Fe/H] = -0.30(0.06) dex for NGC 1193 and [Fe/H]=-0.20(0.07) dex for NGC 1798. The isochrone fitting distance and age of NGC 1193 are 5562(381) pc and 4.6(1) Gyr, respectively. For NGC 1798, these parameters are 4451(728) pc and 1.3(0.2) Gyr. These ages indicate that NGC 1193 and NGC 1798 are old open clusters. The overall present-day mass function slopes for main-sequence stars are found as 1.38(2.16) for NGC 1193 and 1.30(0.21) for NGC 1798, which are in fair agreement with the value of Salpeter (1955). Kinematic and dynamic orbital calculations indicate that NGC 1193 and NGC 1798 belong to the thick-disk and thin-disk populations, respectively. In addition, both of the clusters were born outside the solar circle, and both orbit in the metal-poor region of the Galactic disk.
△ Less
Submitted 13 July, 2022;
originally announced July 2022.
-
Global Optimization via Optimal Decision Trees
Authors:
Dimitris Bertsimas,
Berk Öztürk
Abstract:
The global optimization literature places large emphasis on reducing intractable optimization problems into more tractable structured optimization forms. In order to achieve this goal, many existing methods are restricted to optimization over explicit constraints and objectives that use a subset of possible mathematical primitives. These are limiting in real-world contexts where more general expli…
▽ More
The global optimization literature places large emphasis on reducing intractable optimization problems into more tractable structured optimization forms. In order to achieve this goal, many existing methods are restricted to optimization over explicit constraints and objectives that use a subset of possible mathematical primitives. These are limiting in real-world contexts where more general explicit and black box constraints appear. Leveraging the dramatic speed improvements in mixed-integer optimization (MIO) and recent research in machine learning, we propose a new method to learn MIO-compatible approximations of global optimization problems using optimal decision trees with hyperplanes (OCT-Hs). This constraint learning approach only requires a bounded variable domain, and can address both explicit and inexplicit constraints. We solve the MIO approximation efficiently to find a near-optimal, near-feasible solution to the global optimization problem. We further improve the solution using a series of projected gradient descent iterations. We test the method on a number of numerical benchmarks from the literature as well as real-world design problems, demonstrating its promise in finding global optima efficiently.
△ Less
Submitted 12 February, 2022;
originally announced February 2022.
-
Rastall gravity extension of the standard $Λ$CDM model: theoretical features and observational constraints
Authors:
Ozgur Akarsu,
Nihan Katirci,
Suresh Kumar,
Rafael C. Nunes,
Burcu Ozturk,
Shivani Sharma
Abstract:
We present a detailed investigation of the Rastall gravity extension of the standard $Λ$CDM model. We review the model for two simultaneous modifications of different nature in the Friedmann equation due to the Rastall gravity: the new contributions of the material (actual) sources (considered as effective source) and the altered evolution of the material sources. We discuss the role/behavior of t…
▽ More
We present a detailed investigation of the Rastall gravity extension of the standard $Λ$CDM model. We review the model for two simultaneous modifications of different nature in the Friedmann equation due to the Rastall gravity: the new contributions of the material (actual) sources (considered as effective source) and the altered evolution of the material sources. We discuss the role/behavior of these modifications with regard to some low redshift tensions, including the so-called $H_0$ tension, prevailing within the standard $Λ$CDM. We constrain the model at the level of linear perturbations, and obtain the first constraints through a robust and accurate analysis using the latest full Planck CMB data, with and without including BAO data. We find that the Rastall parameter $ε$ (null for general relativity) is consistent with zero at 68\% CL (with a tendency towards positive values, $-0.0001 < ε< 0.0007$ (CMB+BAO) at 68\% CL), which in turn implies no significant statistical evidence for deviation from general relativity, and also a precision of $\mathcal{O}(10^{-4})$ for the coefficient $-1/2$ of the term $g_{μν}R$ in the Einstein field equations of general relativity (guaranteeing the local energy-momentum conservation). We explore the consequences led by the Rastall gravity on the cosmological parameters in the light of the observational analyses. It turns out that the effective source dynamically screens the usual vacuum energy at high redshifts, but this mechanism barely works due to the opposition by the altered evolution of CDM. Consequently, two simultaneous modifications of different nature in the Friedmann equation act against each other, and do not help to considerably relax the so-called low redshift tensions. Our results may offer a guide for the research community that studies the Rastall gravity in various aspects of gravitation and cosmology.
△ Less
Submitted 15 November, 2020; v1 submitted 8 April, 2020;
originally announced April 2020.
-
A Hybrid Approach to Dependency Parsing: Combining Rules and Morphology with Deep Learning
Authors:
Şaziye Betül Özateş,
Arzucan Özgür,
Tunga Güngör,
Balkız Öztürk
Abstract:
Fully data-driven, deep learning-based models are usually designed as language-independent and have been shown to be successful for many natural language processing tasks. However, when the studied language is low-resourced and the amount of training data is insufficient, these models can benefit from the integration of natural language grammar-based information. We propose two approaches to depen…
▽ More
Fully data-driven, deep learning-based models are usually designed as language-independent and have been shown to be successful for many natural language processing tasks. However, when the studied language is low-resourced and the amount of training data is insufficient, these models can benefit from the integration of natural language grammar-based information. We propose two approaches to dependency parsing especially for languages with restricted amount of training data. Our first approach combines a state-of-the-art deep learning-based parser with a rule-based approach and the second one incorporates morphological information into the parser. In the rule-based approach, the parsing decisions made by the rules are encoded and concatenated with the vector representations of the input words as additional information to the deep network. The morphology-based approach proposes different methods to include the morphological structure of words into the parser network. Experiments are conducted on the IMST-UD Treebank and the results suggest that integration of explicit knowledge about the target language to a neural parser through a rule-based parsing system and morphological analysis leads to more accurate annotations and hence, increases the parsing performance in terms of attachment scores. The proposed methods are developed for Turkish, but can be adapted to other languages as well.
△ Less
Submitted 24 February, 2020;
originally announced February 2020.
