Indian Institute of Technology Palakkad&nbsp;(IIT Palakkad or IIT PKD) is a public autonomous&nbsp;engineering&nbsp;and&nbsp;research&nbsp;institute located in&nbsp;Palakkad,&nbsp;Kerala. It was&nbsp;one of the five new&nbsp;IITs&nbsp;proposed in the&nbsp;2014 Union budget of India. The campus was inaugurated on 3 August 2015. The institute has 94 faculty, 978 students, and 63 non-teaching staffs.

IT Palakkad currently offers four-year Bachelor of Technology (B.Tech) and two years MS programmes in 8 different engineering disciplines including&nbsp;Computer Science, Electrical, Mechanical, Civil among others.

&nbsp;

IIT Palakkad

In this paper, we present a novel unsupervised domain adaptation framework, Multi-Partition Feature Alignment Network, that learns a deep neural model for the target domain without the need for any supervision. Recent leading approaches for unsupervised domain adaptation are based on adversarial alignment. Aligning the global distribution of the domain representations via adversarial training does not guarantee the class-wise distribution alignment. The proposed approach is built on adversarial learning with the focus on carefully aligning class-wise domain representations. Our algorithm utilizes the pseudo-labels (the predicted labels) of the target features to stimulate class-wise alignment. As the pseudo-labels of individual target features can be erroneous, instead of iteratively aligning individual target samples, the proposed framework introduces a generic class-specific multi-partition alignment procedure that enables superior class-discriminative alignment of domain representations. The competitive performance of the proposed framework against state-of-the-art approaches over a wide variety of visual recognition tasks, namely, the digits classification task and the object recognition task, validates its effectiveness for unsupervised domain adaptation.

World University root node

Department of Data Science

Proceedings of the International Joint Conference on Neural Networks

Multi-Partition Feature Alignment Network for Unsupervised Domain Adaptation

Electrical discharge machining (EDM) is one of the nonconventional machining processes suitable for machining of metal matrix composites (MMC). In this work, the heat transfer model has been adopted to predict temperature distribution within selected MMC and considered other machining conditions also such as temperature-dependent thermal properties, Gaussian heat flux and plasma radius. The MMC (Al–4Cu–6Si+10%SiC) is used as a workpiece material in this present investigation. The FEA finite element analysis (FEA) axisymmetric model was generated to simulate the MMC using FEA package multi-physics and predicted Material Removal Rate (MRR). Also, the results have been validated with experimental results. © 2020, Springer Nature Singapore Pte Ltd.

Department of Mechanical Engineering

Advances in Simulation, Product Design and Development

FEA of Electrical Discharge Machining on the Particle Metal Matrix Composite

In the intracellular environment, the intrinsic dynamics of microtubule filaments is often hindered by the presence of barriers of various kind, such as kinetochore complexes and cell cortex, which impact their polymerisation force and dynamical properties such as catastrophe frequency. We present a theoretical study of the effect of a forced barrier, also subjected to thermal noise, on the statistics of catastrophe events in a single microtubule as well as a 'bundle' of two parallel microtubules. For microtubule dynamics, which includes growth, detachment, hydrolysis and the consequent dynamic instability, we employ a one-dimensional discrete stochastic model. The dynamics of the barrier is captured by over-damped Langevin equation, while its interaction with a growing filament is assumed to be hard-core repulsion. A unified treatment of the continuum dynamics of the barrier and the discrete dynamics of the filament is realized using a hybrid Fokker-Planck equation. An explicit mathematical formula for the force-dependent catastrophe frequency of a single microtubule is obtained by solving the above equation, under some assumptions. The prediction agrees well with results of numerical simulations in the appropriate parameter regime. More general situations are studied via numerical simulations. To investigate the extent of 'load-sharing' in a microtubule bundle, and its impact on the frequency of catastrophes, the dynamics of a two-filament bundle is also studied. Here, two parallel, non-interacting microtubules interact with a common, forced barrier. The equations for the two-filament model, when solved using a mean-field assumption, predicts equal sharing of load between the filaments. However, numerical results indicate the existence of a wide spectrum of load-sharing behaviour, which is characterized using a dimensionless parameter. © 2020 IOP Publishing Ltd.

Department of Physics

Physical Biology

Microtubule catastrophe under force: Mathematical and computational results from a Brownian ratchet model

The behavior of droplets of urea-water-solution (UWS) evaporating under the influence of a hot stream of air was investigated experimentally, under temperatures ranging from 100°C to 400°C. The droplets were suspended on a glass microfiber to minimize the influence of heat conduction, through the fiber, on the evaporation rate of the droplet. The flow rate of air, under all experimental conditions, was measured and these data were used to estimate the average velocity of air around the droplet. Experiments were also conducted on droplets of pure water and the results were compared. The initial mass fraction of urea, in the solution, did not appear to have a significant effect on the evaporation constant, but it did affect a few essential aspects of the evaporation behavior. The evaporation of water droplets was in accordance with the d2 law at all temperatures, whereas the evaporation of UWS droplets was ambient-temperature dependent. © 2019 American Institute of Chemical Engineers

AIChE Journal

Experimental investigation of the evaporation behavior of urea‐water‐solution droplets exposed to a hot air stream

The particle/droplet image analysis (PDIA) technique is often used for droplet sizing in sprays. Successful use of PDIA for drop size measurement in sprays mandates careful calibration and determination of the depth of field (DOF) for particles of different size classes. This is typically done with a calibration target that has circles etched on it. In the present study, a calibration experiment was performed with a calibration target as well as droplets, to evaluate the uncertainties in the measured sizes and also to compare the measured DOF values. A continuous-jet droplet generator was used to produce mono-disperse droplets of sizes ranging from 54-393 µm and the calibration glass plate used contained circles of sizes ranging from 20-400 µm. Images of these were captured using a CCD camera equipped with a long-distance microscope and were processed using a code written in MATLAB®. Differences in the measured sizes of droplets and circles and their DOF values are reported. It was observed that the use of the calibration target led to an underestimation of the error in droplet size measurements and an overestimation of the DOF. For the droplets, the measurement error was observed to be less than 5%, with the parameters chosen. © 2020 IOP Publishing Ltd.

Measurement Science and Technology

Estimation of measurement error and depth of field in PDIA experiments: a comparative study using water droplets and a calibration target

Complete characterization of a noisy multipartite quantum state in terms of entanglement requires full knowledge of how the entanglement content in the state is affected by the spatial distribution of noise in the state. Specifically, we find that if the measurement-basis in the protocol of computing localizable entanglement and the basis of the Kraus operator representing the local noisy channel do not commute, the information regarding the noise is retained in the system even after the qubit is traced out after measurement. Using this result and the basic properties of entanglement under noise, we present a set of hierarchies that localizable entanglement over a specific subsystem in a multiqubit state can obey when local noise acts on the subparts or on all the qubits of the whole system. In particular, we propose two types of hierarchies -- one tailored according to the number of noisy unmeasured qubits and the other one that depends additionally on the cardinality of the set of noisy measured qubits, leading to the classification of quantum states. We report the percentage of states satisfying the proposed hierarchies in the case of random three- and four-qubit systems and show, using both analytical methods and numerical simulations, that in almost all the cases, anticipated hierarchies tend to hold with the variation of the strength of noise.

arXiv preprint arXiv:2003.

Hierarchies of localizable entanglement due to spatial distribution of local noise

We demonstrate the existence of universal features in the finite-time thermodynamics of quantum machines by considering a many-body quantum Otto cycle in which the working medium is driven across quantum critical points during the unitary strokes. Specifically, we consider a quantum engine powered by dissipative energizing and relaxing baths. We show that under very generic conditions, the output work is governed by the Kibble-Zurek mechanism; i.e., it exhibits a universal power-law scaling with the driving speed through the critical points. We also optimize the finite-time thermodynamics as a function of the driving speed. The maximum power and the corresponding efficiency take a universal form, and are reached for an optimal speed that is governed by the critical exponents. We exemplify our results by considering a transverse-field Ising spin chain as the working medium. For this model, we also show how the efficiency and power vary as the engine becomes critical. © 2020 authors. Published by the American Physical Society. Published by the American Physical Society under the terms of the Creative Commons Attribution 4.0 International license. Further distribution of this work must maintain attribution to the author(s) and the published article's title, journal citation, and DOI.

Fulltext

Physical Review Research

Universal finite-time thermodynamics of many-body quantum machines from Kibble-Zurek scaling

A graph is called a split graph if its vertex set can be partitioned into a clique and an independent set. Split graphs have rich mathematical structure and interesting algorithmic properties making it one of the most well-studied special graph classes. In the SPLIT VERTEX DELETION problem, given a graph and a positive integer k, the objective is to test whether there exists a subset of at most k vertices whose deletion results in a split graph. In this paper, we design a kernel for this problem with O(k2) vertices, improving upon the previous cubic bound known. Also, by giving a simple reduction from the VERTEX COVER problem, we establish that SPLIT VERTEX DELETION does not admit a kernel with O(k2−ϵ) edges, for any ϵ&gt;0, unless NP⊆coNP/poly. © 2020 Elsevier B.V.