arXiv subjects

Deepak Gupta

Publications and source records attributed to Deepak Gupta.

90 records · Page 5Linked to original sources

arXiv2020

Hierarchical Deep Multi-modal Network for Medical Visual Question Answering

Visual Question Answering in Medical domain (VQA-Med) plays an important role in providing medical assistance to the end-users. These users are expected to raise either a straightforward question with a Yes/No answer or a challenging question that requires a detailed and descriptive answer. The existing techniques in VQA-Med fail to distinguish between the different question types sometimes complicates the simpler problems, or over-simplifies the complicated ones. It is certainly true that for different question types, several distinct systems can lead to confusion and discomfort for the end-users. To address this issue, we propose a hierarchical deep multi-modal network that analyzes and classifies end-user questions/queries and then incorporates a query-specific approach for answer prediction. We refer our proposed approach as Hierarchical Question Segregation based Visual Question Answering, in short HQS-VQA. Our contributions are three-fold, viz. firstly, we propose a question segregation (QS) technique for VQAMed; secondly, we integrate the QS model to the hierarchical deep multi-modal neural network to generate proper answers to the queries related to medical images; and thirdly, we study the impact of QS in Medical-VQA by comparing the performance of the proposed model with QS and a model without QS. We evaluate the performance of our proposed model on two benchmark datasets, viz. RAD and CLEF18. Experimental results show that our proposed HQS-VQA technique outperforms the baseline models with significant margins. We also conduct a detailed quantitative and qualitative analysis of the obtained results and discover potential causes of errors and their solutions.

cs.CL↗

arXiv2020

A Two-Stage Metaheuristic Algorithm for the Dynamic Vehicle Routing Problem in Industry 4.0 approach

Industry 4.0 is a concept that assists companies in developing a modern supply chain (MSC) system when they are faced with a dynamic process. Because Industry 4.0 focuses on mobility and real-time integration, it is a good framework for a dynamic vehicle routing problem (DVRP). This research works on DVRP. The aim of this research is to minimize transportation cost without exceeding the capacity constraint of each vehicle while serving customer demands from a common depot. Meanwhile, new orders arrive at a specific time into the system while the vehicles are executing the delivery of existing orders. This paper presents a two-stage hybrid algorithm for solving the DVRP. In the first stage, construction algorithms are applied to develop the initial route. In the second stage, improvement algorithms are applied. Experimental results were designed for different sizes of problems. Analysis results show the effectiveness of the proposed algorithm.

math.OC↗

arXiv2020

Diagnosis of Breast Cancer Based on Modern Mammography using Hybrid Transfer Learning

Breast cancer is a common cancer for women. Early detection of breast cancer can considerably increase the survival rate of women. This paper mainly focuses on transfer learning process to detect breast cancer. Modified VGG (MVGG), residual network, mobile network is proposed and implemented in this paper. DDSM dataset is used in this paper. Experimental results show that our proposed hybrid transfers learning model (Fusion of MVGG16 and ImageNet) provides an accuracy of 88.3% where the number of epoch is 15. On the other hand, only modified VGG 16 architecture (MVGG 16) provides an accuracy 80.8% and MobileNet provides an accuracy of 77.2%. So, it is clearly stated that the proposed hybrid pre-trained network outperforms well compared to single architecture. This architecture can be considered as an effective tool for the radiologists in order to reduce the false negative and false positive rate. Therefore, the efficiency of mammography analysis will be improved.

eess.IV↗

arXiv2020

Inferring entropy production from short experiments

We provide a strategy for an exact inference of the average as well as the fluctuations of the entropy production in non-equilibrium systems in the steady state, from the measurements of arbitrary current fluctuations. Our results are built upon the finite time generalization of the thermodynamic uncertainty relation, and require only very short time series data from experiments. We illustrate our results with exact and numerical solutions for two colloidal heat engines.

cond-mat.stat-mech↗

arXiv2020

Work fluctuations and Jarzynski equality in stochastic resetting

We consider the paradigm of an overdamped Brownian particle in a potential well, which is modulated through an external protocol, in the presence of stochastic resetting. Thus, in addition to the short range diffusive motion, the particle also experiences intermittent long jumps which reset the particle back at a preferred location. Due to the modulation of the trap, work is done on the system and we investigate the statistical properties of the work fluctuations. We find that the distribution function of the work typically, in asymptotic times, converges to a universal Gaussian form for any protocol as long as that is also renewed after each resetting event. When observed for a finite time, we show that the system does not generically obey the Jarzynski equality which connects the finite time work fluctuations to the difference in free energy, albeit a restricted set of protocols which we identify herein. In stark contrast, the Jarzynski equality is always fulfilled when the protocols continue to evolve without being reset. We present a set of exactly solvable models, demonstrate the validation of our theory and carry out numerical simulations to illustrate these findings. Finally, we have pointed out possible realistic implementations for resetting in experiments using the so-called engineered swift equilibration.

cond-mat.stat-mech↗

arXiv2020

Thermodynamic uncertainty relations in a linear system

We consider a Brownian particle in harmonic confinement of stiffness $k$, in one dimension in the underdamped regime. The whole setup is immersed in a heat bath at temperature $T$. The center of harmonic trap is dragged under any arbitrary protocol. The thermodynamic uncertainty relations for both position of the particle and current at time $t$ are obtained using the second law of thermodynamics as well as the positive semi-definite property of the correlation matrix of work and degrees of freedom of the system for both underdamped and overdamped cases.

cond-mat.stat-mech↗

arXiv2020

Hiding Data in Images Using Cryptography and Deep Neural Network

Steganography is an art of obscuring data inside another quotidian file of similar or varying types. Hiding data has always been of significant importance to digital forensics. Previously, steganography has been combined with cryptography and neural networks separately. Whereas, this research combines steganography, cryptography with the neural networks all together to hide an image inside another container image of the larger or same size. Although the cryptographic technique used is quite simple, but is effective when convoluted with deep neural nets. Other steganography techniques involve hiding data efficiently, but in a uniform pattern which makes it less secure. This method targets both the challenges and make data hiding secure and non-uniform.

cs.MM↗

arXiv2019

Entropy production for partially observed harmonic systems

The probability distribution of the total entropy production in the non-equilibrium steady state follows a symmetry relation called the fluctuation theorem. When a certain part of the system is masked or hidden, it is difficult to infer the exact estimate of the total entropy production. Entropy produced from the observed part of the system shows a significant deviation from the steady state fluctuation theorem. This deviation occurs due to the interaction between the observed and the masked part of the system. A naive guess would be that the deviation from the steady state fluctuation theorem may disappear in the limit of small interaction between both parts of the system. In contrast, we investigate the entropy production of a particle in a harmonically coupled Brownian particle system (say, particle A and B) in a heat reservoir at a constant temperature. The system is maintained in the non-equilibrium steady state using stochastic driving. When the coupling between particle A and B is infinitesimally weak, the deviation from the steady state fluctuation theorem for the entropy production of a partial system of a coupled system is studied. Furthermore, we consider a harmonically confined system (i.e., a harmonically coupled system of particle A and B in harmonic confinement). In the weak coupling limit, the entropy produced by the partial system (e.g., particle A) of the coupled system in a harmonic trap satisfies the steady state fluctuation theorem. Numerical simulations are performed to support the analytical results. Part of these results was announced in a recent letter, Europhys. Lett. 115, 60003 (2016).

cond-mat.stat-mech↗

arXiv2019

Shall we turn off the media? Global information can destroy local cooperation in the one-dimensional ring

In this paper we investigate the evolution of cooperation when the interaction structure is strictly local, and hence fitness only depends on local behaviors, while the competition structure is partly global, and hence selection can happen also between distant agents. We explore this novel setup by means of numerical simulations in a model where agents are arranged in a one-dimensional ring. Preliminary results suggest that the extent of global comparison systematically favors defection under strong selection, while its effect under weak selection is less systematic, but overall still favors defection. Further, the extent of global comparison seems to reduce sensibly fixation times.

physics.soc-ph↗

arXiv2018

Stochastic resetting in underdamped Brownian motion

We consider a single Brownian particle in one dimension in a medium at a constant temperature in the underdamped regime. We stochastically reset the position of the Brownian particle to a fixed point in the space with a constant rate $r$ whereas its velocity evolves irrespective of the position of the particle and stochastic resetting mechanism. The nonequilibrium steady state of the position distribution is studied for this model system. Further, we study the distribution of the position of the particle in the finite time and the approach to the nonequilibrium steady state distribution with time. Numerical simulations are done to verify the analytical results.

cond-mat.stat-mech↗

arXiv2018

Helping each Other: A Framework for Customer-to-Customer Suggestion Mining using a Semi-supervised Deep Neural Network

Suggestion mining is increasingly becoming an important task along with sentiment analysis. In today's cyberspace world, people not only express their sentiments and dispositions towards some entities or services, but they also spend considerable time sharing their experiences and advice to fellow customers and the product/service providers with two-fold agenda: helping fellow customers who are likely to share a similar experience, and motivating the producer to bring specific changes in their offerings which would be more appreciated by the customers. In our current work, we propose a hybrid deep learning model to identify whether a review text contains any suggestion. The model employs semi-supervised learning to leverage the useful information from the large amount of unlabeled data. We evaluate the performance of our proposed model on a benchmark customer review dataset, comprising of the reviews of Hotel and Electronics domains. Our proposed approach shows the F-scores of 65.6% and 65.5% for the Hotel and Electronics review datasets, respectively. These performances are significantly better compared to the existing state-of-the-art system.

cs.CL↗

arXiv2018

Combining Graph-based Dependency Features with Convolutional Neural Network for Answer Triggering

Answer triggering is the task of selecting the best-suited answer for a given question from a set of candidate answers if exists. In this paper, we present a hybrid deep learning model for answer triggering, which combines several dependency graph based alignment features, namely graph edit distance, graph-based similarity and dependency graph coverage, with dense vector embeddings from a Convolutional Neural Network (CNN). Our experiments on the WikiQA dataset show that such a combination can more accurately trigger a candidate answer compared to the previous state-of-the-art models. Comparative study on WikiQA dataset shows 5.86% absolute F-score improvement at the question level.

cs.AI↗

arXiv2018

Exact distribution for work and stochastic efficiency of an isothermal machine

We consider an isothermal machine composed of two Brownian particles (say particle A and B) connected by a harmonic spring. A constant load is attached to particle A, and the particle B is trapped in a harmonic confinement whose minimum is dragged with a constant velocity. Whole system is in contact with the heat bath of a constant temperature. We obtain the distribution of the work done on particle A and particle B, and transient fluctuation theorem for these quantities is tested in the weak coupling limit and for both small and large observation time. Moreover, we show that the transient fluctuation theorem for total work done on both particles is satisfied. Furthermore, we compute the stochastic efficiency which is the ratio of the work done against the load force on particle A and the work done on particle B of this machine. The probability density function for stochastic efficiency is computed for all time. Numerical simulations are also done to verify the analytical results.

cond-mat.stat-mech↗

arXiv2017

Partial entropy production in heat transport

We consider a system of two Brownian particles (say A and B), coupled to each other via harmonic potential of stiffness constant $k$. Particle-A is connected to two heat baths of constant temperatures $T_1$ and $T_2$, and particle-B is connected to a single heat bath of a constant temperature $T_3$. In the steady state, the total entropy production for both particles obeys the fluctuation theorem. We compute the total entropy production due to one of the particles called as partial or apparent entropy production, in the steady state for a time segment $τ$. When both particles are weakly interacting with each other, the fluctuation theorem for partial and apparent entropy production is studied. We find a significant deviation from the fluctuation theorem. The analytical results are also verified using numerical simulations.

cond-mat.stat-mech↗

arXiv2017

Auto Analysis of Customer Feedback using CNN and GRU Network

Analyzing customer feedback is the best way to channelize the data into new marketing strategies that benefit entrepreneurs as well as customers. Therefore an automated system which can analyze the customer behavior is in great demand. Users may write feedbacks in any language, and hence mining appropriate information often becomes intractable. Especially in a traditional feature-based supervised model, it is difficult to build a generic system as one has to understand the concerned language for finding the relevant features. In order to overcome this, we propose deep Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN) based approaches that do not require handcrafting of features. We evaluate these techniques for analyzing customer feedback sentences in four languages, namely English, French, Japanese and Spanish. Our empirical analysis shows that our models perform well in all the four languages on the setups of IJCNLP Shared Task on Customer Feedback Analysis. Our model achieved the second rank in French, with an accuracy of 71.75% and third ranks for all the other languages.

cs.CL↗

arXiv2017

SMPOST: Parts of Speech Tagger for Code-Mixed Indic Social Media Text

Use of social media has grown dramatically during the last few years. Users follow informal languages in communicating through social media. The language of communication is often mixed in nature, where people transcribe their regional language with English and this technique is found to be extremely popular. Natural language processing (NLP) aims to infer the information from these text where Part-of-Speech (PoS) tagging plays an important role in getting the prosody of the written text. For the task of PoS tagging on Code-Mixed Indian Social Media Text, we develop a supervised system based on Conditional Random Field classifier. In order to tackle the problem effectively, we have focused on extracting rich linguistic features. We participate in three different language pairs, ie. English-Hindi, English-Bengali and English-Telugu on three different social media platforms, Twitter, Facebook & WhatsApp. The proposed system is able to successfully assign coarse as well as fine-grained PoS tag labels for a given a code-mixed sentence. Experiments show that our system is quite generic that shows encouraging performance levels on all the three language pairs in all the domains.

cs.CL↗

arXiv2016

Fluctuation theorem for entropy production of a partial system in the weak coupling limit

Small systems in contact with a heat bath evolve by stochastic dynamics. Here we show that, when one such small system is weakly coupled to another one, it is possible to infer the presence of such weak coupling by observing the violation of the steady state fluctuation theorem for the partial entropy production of the observed system. We give a general mechanism due to which the violation of the fluctuation theorem can be significant, even for weak coupling. We analytically demonstrate on a realistic model system that this mechanism can be realized by applying an external random force to the system. In other words, we find a new fluctuation theorem for the entropy production of a partial system, in the limit of weak coupling.

cond-mat.stat-mech↗

arXiv2011

Throughput Performance of 2$\times$2 Mimo LTE Downlink in a Spatial Correlation Based Microcellular Channel for Wireless Broadband Networks

Multiple Input Multiple Output (MIMO) technology is going to be a viable alternative for future generation wireless broadband services in order to meet the striving requirements for throughput and system robustness. In Long Term Evolution (LTE), MIMO technologies have been broadly used to get better downlink peak rate, cell coverage, as well as average cell throughput. In the present paper a 2x2 MIMO is taken as baseline configuration for a LTE downlink under a Microcellular propagation scenario considering a non physical correlation based channel with Poor and rich scattering environment. The throughput capacity of the downlink is obtained for poor and rich scattering environments. Besides, two vital aspects of MIMO technique viz Spatial Multiplexing (SM) and Transmit Diversity (TD) are investigated in order to see their effect on throughput of the system. The effect of parameters like Speed of mobile station, number of Multipath, Rician factor (K) on throughput of such systems is reported and discussed. The investigations reported in this paper helps in estimating the throughput capacity of LTE downlink under SM and TD mode.

cs.NI↗