arXiv subjects

Amandeep Kaur

Publications and source records attributed to Amandeep Kaur.

17 recordsLinked to original sources

arXiv2026

Learning to replenish: A hybrid deep reinforcement learning for dynamic inventory management in the pharmaceutical supply chains

Pharmaceutical supply chains (PSCs) struggle with inventory management (IM) due to unpredictable demand patterns and variable lead times associated with restocking. This complexity is further compounded by the finite shelf lives of pharmaceutical products, which necessitate a delicate balance between adequate stock and minimal waste. These intertwined factors create a complex optimization problem that requires sophisticated inventory strategies to ensure both product availability and PSC efficiency. This study aims to develop an optimal inventory replenishment policy for pharmaceutical products that can handle the stochasticity arising from uncertain demand and variable PSC conditions. The objective is to maximize the profitability of the PSC while maintaining a high patient service level. We formulate the problem as a Markov decision process and propose a deep reinforcement learning (DRL) approach, specifically, a hybrid asynchronous advantage actor critic distributed proximal policy optimization (A3C DPPO)algorithm. The A3C DPPO algorithm is tailored to handle the continuous action space inherent in IM. The numerical results demonstrate that the proposed algorithm adaptively updates the inventory replenishment strategy under dynamic scenarios, resulting in lower inventory costs compared to various benchmarks. We also conduct numerical validation using real-world pharmaceutical inventory data to confirm the practical feasibility of the proposed algorithm.

cs.AI↗

arXiv2026

Pretrain Where? Investigating How Pretraining Data Diversity Impacts Geospatial Foundation Model Performance

New geospatial foundation models introduce a new model architecture and pretraining dataset, often sampled using different notions of data diversity. Performance differences are largely attributed to the model architecture or input modalities, while the role of the pretraining dataset is rarely studied. To address this research gap, we conducted a systematic study on how the geographic composition of pretraining data affects a model's downstream performance. We created global and per-continent pretraining datasets and evaluated them on global and per-continent downstream datasets. We found that the pretraining dataset from Europe outperformed global and continent-specific pretraining datasets on both global and local downstream evaluations. To investigate the factors influencing a pretraining dataset's downstream performance, we analysed 10 pretraining datasets using diversity across continents, biomes, landcover and spectral values. We found that only spectral diversity was strongly correlated with performance, while others were weakly correlated. This finding establishes a new dimension of diversity to be accounted for when creating a high-performing pretraining dataset. We open-sourced 7 new pretraining datasets, pretrained models, and our experimental framework at https://github.com/kerner-lab/pretrain-where.

cs.CV↗

arXiv2026

Si-Ga2O3/p-GaN epitaxial heterostructure based self-powered and visible-blind UV photodetectors with fast and electrically tuneable response time

n-Ga2O3/p-GaN heterojunction based photodetector devices are fabricated on Si-doped (-201) β-Ga2O3 epitaxial layers grown by pulsed laser deposition (PLD) technique on p-type c-GaN/sapphire templates. These devices demonstrate the ability to act as highly efficient self-powered visible blind UV-photodetectors with fast response time. It has been found that the optimum performance of the detector in terms of its responsivity, detectivity and response time could be achieved by adjusting the Si doping level and the thickness of the Ga2O3 layer. Our best performing device showing the peak responsivity and detectivity of 56.8 mA/W and 3*10^12 Jones, respectively, is achieved for 660 nm thick Ga2O3 layer with Si-concentration of 8*10^18 cm^-3. Moreover, as low as a few nW of optical signal can be sensed by the detector. The response time of the detector is found to be only a few tens of nanoseconds, which highlights their potential for application in ultrafast detection of UV light. These devices also exhibit a slower component of photoresponse with a timescale of a few tens of milliseconds. Interestingly, the time-scale of the slower response can be prolongated by several orders of magnitude through enhancing the applied reverse bias. Such an electrical tuneability of the response time is highly desirable for neuromorphic device applications.

cond-mat.mtrl-sci↗

arXiv2026

CVD grown bilayer MoS2 based artificial optoelectronic synapses for arithmetic computing and image recognition applications

Demand for lower computing power has rapidly increased. In this context, brain-inspired neuromorphic computing, which integrate data storage and processing, has attracted significant attention. Here, our study reveals that field effect transistors fabricated on chemical vapor deposited bilayer (2L) MoS2 films can mimic the functions of biological synapse. These devices demonstrate high level of pair pulse facilitation (PPF), short term to long term memory (STM-to-LTM) transition as well as learning-forgetting-relearning properties. Effect of light intensity, pulse number, pulse width and photon energy on the STM-to-LTM transition is studied. It has been found that the rate of depression of the memory state can be controlled using the gate bias. Electrical and optical energy consumptions per synaptic event are estimated to be as low as 280 fJ and 20 nJ, respectively. Furthermore, photocurrent in these devices is observed to increase linearly with the number of the excitation pulses. This property has been exploited to demonstrate different arithmetic operations by the device. Moreover, these devices show great potential for image recognition. Artificial neural network simulation has returned an image recognition accuracy of ~85%. All these findings show a great prospect of 2L-MoS2 for developing low power, transparent and flexible neuromorphic devices.

cond-mat.mtrl-sci↗

arXiv2025

Adaptive Inventory Strategies using Deep Reinforcement Learning for Dynamic Agri-Food Supply Chains

Agricultural products are often subject to seasonal fluctuations in production and demand. Predicting and managing inventory levels in response to these variations can be challenging, leading to either excess inventory or stockouts. Additionally, the coordination among stakeholders at various level of food supply chain is not considered in the existing body of literature. To bridge these research gaps, this study focuses on inventory management of agri-food products under demand and lead time uncertainties. By implementing effective inventory replenishment policy results in maximize the overall profit throughout the supply chain. However, the complexity of the problem increases due to these uncertainties and shelf-life of the product, that makes challenging to implement traditional approaches to generate optimal set of solutions. Thus, the current study propose a novel Deep Reinforcement Learning (DRL) algorithm that combines the benefits of both value- and policy-based DRL approaches for inventory optimization under uncertainties. The proposed algorithm can incentivize collaboration among stakeholders by aligning their interests and objectives through shared optimization goal of maximizing profitability along the agri-food supply chain while considering perishability, and uncertainty simultaneously. By selecting optimal order quantities with continuous action space, the proposed algorithm effectively addresses the inventory optimization challenges. To rigorously evaluate this algorithm, the empirical data from fresh agricultural products supply chain inventory is considered. Experimental results corroborate the improved performance of the proposed inventory replenishment policy under stochastic demand patterns and lead time scenarios. The research findings hold managerial implications for policymakers to manage the inventory of agricultural products more effectively under uncertainty.

cs.AI↗

arXiv2025

Towards AI-Driven RANs for 6G and Beyond: Architectural Advancements and Future Horizons

It is envisioned that 6G networks will be supported by key architectural principles, including intelligence, decentralization, interoperability, and digitalization. With the advances in artificial intelligence (AI) and machine learning (ML), embedding intelligence into the foundation of wireless communication systems is recognized as essential for 6G and beyond. Existing radio access network (RAN) architectures struggle to meet the ever growing demands for flexibility, automation, and adaptability required to build self-evolving and autonomous wireless networks. In this context, this paper explores the transition towards AI-driven RAN (AI-RAN) by developing a novel AI-RAN framework whose performance is evaluated through a practical scenario focused on intelligent orchestration and resource optimization. Besides, the paper reviews the evolution of RAN architectures and sheds light on key enablers of AI-RAN including digital twins (DTs), intelligent reflecting surfaces (IRSs), large generative AI (GenAI) models, and blockchain (BC). Furthermore, it discusses the deployment challenges of AI-RAN, including technical and regulatory perspectives, and outlines future research directions incorporating technologies such as integrated sensing and communication (ISAC) and agentic AI.

eess.SP↗

arXiv2025

Hot pygmy dipole strength in nickel isotopes

At finite temperatures, nuclear excitations are significantly modified, most notably through the emergence of additional low-energy dipole strength, which can critically impact astrophysical reaction rates. Ongoing fusion-evaporation experiments on Ni isotopes provide a unique opportunity to investigate the hot pygmy dipole strength (HPDS), underscoring the need for reliable theoretical predictions and a comprehensive understanding of this emerging phenomenon. In this work, the HPDS is investigated in Ni isotopes from $N = Z$ to neutron-rich systems ($^{56\text{--}70}$Ni) over a temperature range of $T=$ 0$-$2~MeV using the finite-temperature relativistic quasiparticle random phase approximation. In neutron-rich Ni isotopes, the pygmy dipole strength at higher temperatures exceeds up to 2.5 times its value observed at zero temperature. In contrast, near $N \approx Z$ isotopes show negligible low-energy dipole strength at $T = 0$ MeV but develop a pronounced HPDS as the temperature increases. Predicted E1 energy-weighted strength ($S_{\text{EWS}}$) and cumulative $B$(E1) values for HPDS are presented across the Ni isotopic chain for various low-energy intervals and temperatures, providing essential benchmarks to support and guide experimental studies.

nucl-th↗

arXiv2025

Electric and magnetic $γ$-ray strength functions at finite-temperature

The $γ$-ray strength function ($γ$SF) is essential for understanding the electromagnetic response in atomic nuclei and modeling astrophysical neutron capture rates. We introduced a microscopic description of both electric dipole (E1) and magnetic dipole (M1) $γ$SFs that includes finite-temperature effects within relativistic density functional theory. The temperature dependence of the total electromagnetic $γ$SFs shows significant modification in the low-energy region due to thermal unblocking effects, essential for agreement with recent particle-$γ$ coincidence data from the Oslo method. An investigation of the electric and magnetic contributions to the total $γ$SF in hot nuclei indicates that the M1 mode becomes more prominent in the low-energy region, different than what is known at zero temperature. This microscopic approach offers new insights into the interplay between E1 and M1 $γ$SFs at finite-temperature, and opens new perspectives for future studies of $(n,γ)$ reactions and nucleosynthesis in hot stellar environments.

nucl-th↗

arXiv2024

Moving Beyond LDA: A Comparison of Unsupervised Topic Modelling Techniques for Qualitative Data Analysis of Online Communities

Social media constitutes a rich and influential source of information for qualitative researchers. Although computational techniques like topic modelling assist with managing the volume and diversity of social media content, qualitative researcher's lack of programming expertise creates a significant barrier to their adoption. In this paper we explore how BERTopic, an advanced Large Language Model (LLM)-based topic modelling technique, can support qualitative data analysis of social media. We conducted interviews and hands-on evaluations in which qualitative researchers compared topics from three modelling techniques: LDA, NMF, and BERTopic. BERTopic was favoured by 8 of 12 participants for its ability to provide detailed, coherent clusters for deeper understanding and actionable insights. Participants also prioritised topic relevance, logical organisation, and the capacity to reveal unexpected relationships within the data. Our findings underscore the potential of LLM-based techniques for supporting qualitative analysis.

cs.HC↗

arXiv2024

p-(001)NiO/n-(0001)ZnO Heterostructures based Ultraviolet Photodetectors

We investigate the potential of epitaxial (001)p-NiO/(0001)n-ZnO heterostructures grown on (0001)sapphire substrates by pulsed laser deposition technique for ultraviolet photodetector application. Our study reveals that in the self-powered mode, these devices can serve as effective photodetectors for the UV-A band (320-400 nm) with response time as short as 400 microseconds. Peak responsivity as high as 5mA/W at zero bias condition have been achieved. These devices also show a very high level of stability under repeated on/off illumination cycles over a long period of time. Furthermore, we find that the response time of these detectors can be controlled from several microseconds to thousands of seconds by applying bias both in the forward and the reverse directions. This persistent photoconductivity effect has been explained in terms of the field induced change in the capture barrier height associated with certain traps located at the junction.

physics.app-ph↗

arXiv2024

p-(001)NiO/n-(0001)ZnO heterostructures grown by pulsed laser deposition technique

NiO/ZnO heterostructures are grown on c-sapphire substrates using pulsed laser deposition (PLD) technique. X-ray diffraction study shows that the ZnO layer epitaxially grows along [0001]-direction on (0001)sapphire surface as expected. While, the epitaxial NiO film is found to be deposited along [001]-direction on the (0001)ZnO surface. Moreover, the presence of three (001)NiO domains laterally rotated by 30° with respect to each other, has also been observed in our NiO films. The study reveals the continuous nature of the NiO film, which also possesses a very smooth surface morphology. In a sharp contrast, ZnO films are found to grow along [0001]-direction when deposited on (111)NiO layers. These films also show columnar morphology. (001)NiO/(0001)ZnO layers exhibit the rectifying current-voltage characteristics that suggests the existence of p-n junction in these devices. However, the behavior could not be observed in (0001)ZnO/(111)NiO heterojunctions. The reason could be the columnar morphology of the ZnO layer. Such a morphology can facilitate the propagation of the metal ions from the contact pads to the underlying NiO layer and suppress the p-n junction effect.

cond-mat.mtrl-sci↗

arXiv2023

Finite temperature effects in magnetic dipole transitions

Finite temperature effects in electromagnetic transitions in nuclei contribute to many aspects of nuclear structure and astrophysically relevant nuclear reactions. While electric dipole transitions have already been extensively studied, the temperature sensitivity of magnetic transitions remains largely unknown. This work comprises the study of isovector magnetic dipole excitations (M1) occurring between spin-orbit (SO) partner states using the recently developed self-consistent finite temperature relativistic quasiparticle random phase approximation (FT-RQRPA) in the temperature range from $T=$ 0 to 2 MeV. The M1 strength distributions of $^{40-60}$Ca and $^{100-140}$Sn isotopic chains exhibit a strong temperature dependence. The M1 strength peaks shift significantly towards the lower energies due to the decrease in SO splitting energies and weakening of the residual interaction, especially above the critical temperatures ($T_\textrm{c}$) where the pairing correlations vanish. By exploring the relevant two-quasiparticle ($2qp$) configurations contributing to the M1 strength of closed- and open-shell nuclei, new proton and neutron excitation channels between SO partners are observed in low- and high-energy regions due to the thermal unblocking effects around the Fermi level. At higher temperatures, we have noticed an interesting result in $^{40,60}$Ca nuclei, the appearance of M1 excitations, which are forbidden at zero temperature due to fully occupied (or fully vacant) spin-orbit partner states.

nucl-th↗

arXiv2023

Electric dipole transitions in the relativistic quasiparticle random phase approximation at finite temperature

Finite temperature results in various effects on the properties of nuclear structure and excitations of relevance for nuclear processes in hot stellar environments. Here we introduce the self-consistent finite temperature relativistic quasiparticle random phase approximation (FT-RQRPA) based on relativistic energy density functional with point coupling interaction for describing the temperature effects in electric dipole (E1) transitions. We perform a study of E1 excitations in the temperature range $T=$ 0-2 MeV for the selected closed- and open-shell nuclei ranging from $^{40}$Ca to $^{60}$Ca and $^{100}$Sn to $^{140}$Sn by including both thermal and pairing effects. The isovector giant dipole resonance strength is slightly modified for the considered range of temperature, while new low-energy peaks emerge for $E<$12 MeV with non-negligible strength in neutron-rich nuclei at high temperatures. The analysis of relevant two-quasiparticle configurations discloses how new excitation channels open due to thermal unblocking of states at finite temperature. The study also examines the isospin and temperature dependence of electric dipole polarizability ($α_D$), resulting in systematic increase in the values of $α_D$ with increasing temperature, with a more pronounced effect observed in neutron-rich nuclei. The FT-RQRPA introduced in this work will open perspectives for microscopic calculation of $γ$-ray strength functions at finite temperatures relevant for nuclear reaction studies.

nucl-th↗

arXiv2023

Ni cluster embedded (111)NiO layers grown on (0001)GaN films using pulsed laser deposition technique

(111) NiO epitaxial layers embedded with crystallographically oriented Ni-clusters are grown on c-GaN/Sapphire templates using pulsed laser deposition technique. Structural and magnetic properties of the films are examined by a variety of techniques including high resolution x-ray diffraction, precession-electron diffraction and superconducting quantum interference device magnetometry. The study reveals that the inclusion, orientation, shape, size, density and magnetic properties of these clusters depend strongly on the growth temperature (TG). Though, most of the Ni-clusters are found to be crystallographically aligned with the NiO matrix with Ni(111) parallel to NiO(111), clusters with other orientations also exist, especially in samples grown at lower temperatures. Average size and density of the clusters increase with TG . Proportion of the Ni(111) parallel to NiO(111) oriented clusters also improves as TG is increased. All cluster embedded films show ferromagnetic behaviour even at room temperature. Easy-axis is found to be oriented in the layer plane in samples grown at relatively lower temperatures. However, it turns perpendicular to the layer plane for samples grown at sufficiently high temperatures. This reversal of easy-axis has been attributed to the size dependent competition between the shape, magnetoelastic and the surface anisotropies of the clusters. This composite material thus has great potential to serve as spin-injector and spinstorage medium in GaN based spintronics of the future.

cond-mat.mtrl-sci↗

arXiv2022

Photosensitive SrMnO3

In recent years, photosensitive materials have been in huge demand because of their fascinating ability to convert absorbed photon energy to generate strain and henceforth tuning the physical properties. In this report we detect the photosensitive activity of SrMnO3. Using the power dependent and temperature dependent Raman study with different laser sources having wavelengths across the optical band gap of SrMnO3, we divulge the photosensitive character of SrMnO3 thin films. Upon laser light illumination, Raman modes soften and softening further increases with increase in laser power. Similar kind of mode variation is observed with increasing temperature at fixed laser power. XAS in presence of laser illumination, reveals the change in crystal field splitting associated with mode softening.

cond-mat.mtrl-sci↗

arXiv2020

A comparison of deep machine learning algorithms in COVID-19 disease diagnosis

The aim of the work is to use deep neural network models for solving the problem of image recognition. These days, every human being is threatened by a harmful coronavirus disease, also called COVID-19 disease. The spread of coronavirus affects the economy of many countries in the world. To find COVID-19 patients early is very essential to avoid the spread and harm to society. Pathological tests and Chromatography(CT) scans are helpful for the diagnosis of COVID-19. However, these tests are having drawbacks such as a large number of false positives, and cost of these tests are so expensive. Hence, it requires finding an easy, accurate, and less expensive way for the detection of the harmful COVID-19 disease. Chest-x-ray can be useful for the detection of this disease. Therefore, in this work chest, x-ray images are used for the diagnosis of suspected COVID-19 patients using modern machine learning techniques. The analysis of the results is carried out and conclusions are made about the effectiveness of deep machine learning algorithms in image recognition problems.

cs.LG↗

arXiv2010

First-Principles Study of Electronic and Vibrational Properties of BaHfN$_2$

The transition metal nitride BaHfN$_2$, which consists of weakly bonded neutral slabs of closed shell ions, has structural and chemical similarities to other layered nitrides which have impressive superconducting T$_c$ when electron doped: A$_x$HfNCl, A$_x$ZrNCl, A$_x$TiNCl, with $T_c= 25.5$, $15.2$ and $16.5$ K, respectively for appropriate donor (A) concentrations $x$. These similarities suggest the possibility of BaHfN$_2$ being another relatively high T$_c$ nitride upon doping, with effects of structure and the role of specific transition metal ions yet to be understood. We report first-principles electronic structure calculations for stoichiometric BaHfN$_2$ using density functional theory with plane-wave basis sets and separable dual-space Gaussian pseudopotentials. An indirect band gap of 0.8 eV was obtained and the lowest conduction band is primarily of Hf 5$d_{xy}$ character, similar to $β$-ZrNCl and $α$-TiNCl. The two N sites, one in the Hf layer and another one in the Ba layer, were found to have very anisotropic Born effective charges (BEC):deviations from the formal charge (-3) are opposite for the two sites, and opposite for the two orientations (in-plane, out of plane). LO-TO splittings and comparison of BECs and dielectric constant tensors to those of related compounds are discussed, and the effect of electron doping on the zone-center phonons is reported.

cond-mat.mtrl-sci↗