If the cyclic sequence of faces for all the vertices in a map are of the same types, then the map is a semi-equivelar map. In particular, a semi-equivelar is equivelar if the faces are the same type. Topological quantum codes are being introduced as an alternative quantum code. The homological quantum code is a subclass of topological quantum code. We produce a technique to construct homological quantum codes associate with semi-equivelar maps. We also present homological quantum codes associate with maps on the surfaces with Euler characteristic -1 , -2 . Furthermore, we will present fifteen classes of homological quantum codes associate with covering maps of the class of maps on the surface with Euler characteristic -1 and -2 .
Although the first SHA-1 collision attack has been carried out in 2017, the cryptographic hash function seems still robust against pre-image attacks. The aim of this work is to measure how many rounds of SHA-1 can be inverted in practice using SAT solvers. To do so, we have first modeled the problem of finding a pre-image of SHA-1 as a system of Boolean formulas, explicitly describing the procedure to obtain such model. Then, we configured the model based on different combinations of number of rounds and number of free bits in our target pre-image. Finally, to find a solution of the model, we used a SAT solver on our server, testing several combinations of restart policies and polarity modes. We analyze and report the number and positions of the pre-image bits that can be fixed to influence the ability of the SAT solver to find the remaining free bits of the pre-image in a shorter time. In particular, we execute partial pre-image attacks on 64-, 80- 96-, 112- and 128-bit messages, outperforming the current state-of-the-art records.
Nowadays, mortality due to corona virus (COVID-19) is uttermost important for surveillance and monitoring purpose. It is an illness caused by a virus which spreads from person to person throughout the World. The delay in reporting for such case purpose between the time of occurrence of the mortality and them being registered is unavoidable. This work proposes an approach for estimating the unknown surveillance in a timely manner, based on weekly reporting on mortality. It is also performed that the demographic characteristics such as gender, age, and various diseases responsible for death are significantly associated with the probability of the enlisting report of USA and also predict the influence of the variables. The current work can generate timely and accurate estimations on the weekly count of mortality surveillance only 1-week lag. The proposed technique has executed average measurements with threefold nested cross validation and among all splitting methods the best performance has high coefficients of determination, i.e., 0.99 with MAE (mean absolute error) 63.78 and RMSE (root mean squared error) 93.20. Simulation is based on R-software.
Let T be a spanning subgraph of a finite undirected graph G. T is called a spanning maximal planar subgraph of G if it is also a maximal planar graph, that is, each of its regions is bounded by three edges. In this paper, we focus our attention on some classes of complete 4-partite graphs and determine when these graphs contain the spanning maximal planar subgraph T. In particular, we present some results on the spanning maximal planar subgraph problem for complete 4-partite graphs of the form $$K_{1,1,1,z}$$ and $$K_{1,1,y,z}$$ and utilize these results to attack the problem for general $$K_{w,x,y,z}$$ given some initial conditions for y and z.
Development of new drugs has the limitations of high cost, high time requirement and low success rate. If an existing drug can be used to treat a drug-seeking disease, we can reduce these limitations. The process of using existing drugs to treat new diseases is called drug repurposing. Since the drugs are already approved for human use, the success rate becomes high. In recent years, numerous random walk models have been proposed on the disease-drug heterogeneous network and become a popular drug repurposing approach. The performance of random walk-based approach depends on the network similarity measures used to build the heterogeneous network. In this paper, we improve the network similarity measures by integrating the similarity between the disease and drug-specific protein interactomes in human Protein-Protein Interaction network. We then run a random walk with restart algorithm over the modified network to predict disease-drug relations. Our experiments reveal that performance of random walk model has improved after integrating protein interactome similarity.
Presence of irrelevant sentences containing irrelevant quantities in Math Word Problems (MWPs) throws an immense challenge in formulating the final equation(s). Failure in identification of the relevant quantities with respect to the question asked in an MWP reduces the overall system performance. This paper demonstrates a novel deep learning-based approach using Siamese neural network to classify the relevant and irrelevant sentences towards identification of the relevant quantities in arithmetic MWPs. The proposed relevance classifier produced an accuracy of 91.0
Semi-Markov decision processes with vector rewards are studied here under discounted pay-off criterion. The notion of optimality is replaced by Pareto optimality here. We show that a pure and stationary Pareto-optimal strategy exists along with the existence of an algorithm to construct an approximate version of Pareto curves, i.e., ϵ -approximate Pareto curve in polynomial time as in a multi-objective linear programming problem. Semi-Markov decision processes with multiple objectives find applications in situations like dynamic goal programming where the decision-maker has more than one objective to be optimized simultaneously. Further, we also investigate the Pareto-realizability problem as well as the NP-completeness of pure stationary Pareto-realizability problem.
Predicting the correct values of stocks in fast fluctuating high-frequency financial data is always a challenging task. Existing state-of-the-art models are very efficient in terms of accuracy but lags in prediction speed. In this work, we aim to develop a deep-learning-based fast model for live predictions of stock values with minimum errors. The proposed model is based on fast recurrent neural networks (FastRNNs), which provides us with both of the desired features. We have considered the 1-min time interval stock data of four companies for a period of one day. The model is aimed to have a low computational complexity as well so that it can be run for live predictions as well. The model’s performance is measured by root mean square error (RMSE) along with computation time. The model outperforms LSTM, CNN, and other deep learning models for live predictions of stock values.
In this paper, we outline a specific problem concerning the use of multimodal chaotic maps in the core compression-construction of cryptographic hash functions. In cryptographic hash applications, the security level is measured according to their collision resistance property. This paper analyzes the strength of this property, by showing how it could be affected in the case where message units are modulated directly into the chaotic phase space of a multimodal map. Mathematical properties of multimodal maps are discussed and used to prove the risk of collision existence in such hash schemes. Four examples of widely used chaotic-maps in hash constructions are given and analyzed. Finally, we demonstrate the risk of collision existence using a practical example inherited from a previously proposed hash function.
The well-known Niederreiter encryption scheme is an improved version of the classical McEliece encryption scheme in terms of key size and encryption efficiency, but the conversion of the scheme to a usable signature scheme is a very challenging process. Moreover, the current blockchain architecture still relies on hard problems that cannot resist quantum algorithms attacks. In this paper, we construct a Niederreiter-based signature scheme with a comprehensive security analysis that is suitable to be used together with the blockchain technology and resistance to Shor’s algorithm for better underlying performance and security of the blockchain architecture.
Cloud computing has taken the computing services market by storm due to its advantages over traditional computing platforms. Cloud computing makes its resources available over the Internet, so clients can purchase them and pay only for what is used similar to other utilities. For cloud computing to become acceptable to a wider business clientele including the ones with high expectations, it is vital to maintain the QoS commitments. Hence, it is necessary to have mechanisms and tools to measure and quantify the performance of cloud service providers. In this paper, the authors propose a mathematical model that could measure multiple QoS attributes, quantify them and compute a single score called the trust score to help customers identify the most suitable service provider. The proposed enhanced QoS (eQoS) trust model was tested using simulations. The results show that the proposed mechanism performs better than the other two QoS-based trust models, namely QoS Trust and Turnaround Trust published in the literature.
The employment of data mining, granular knowledge exploration and machine learning approaches in the medical field (several of diseases in bioinformatics) are proven to be prolific as such approaches are thought of computationally empirical within the higher cognitive process of medical practitioners. As an intelligent healthcare strategy, this paper presents a fast and accurate framework for breast cancer disease classification and predictive analysis. In an experimental simulation, we obtained ceiling improvement on prediction accuracy using Unary KNN and Binary LightGBM Stacked Ensemble Learning. We have given the visualization of results and judged the obtained various statistical performance measures. Our empirical evaluation and experimental analysis reflect that this strategy with provable performance guarantees performs well in comparison with other commonly used significant state-of-the-art methods. We observed that the proposed framework can be a powerful tool for simplifying or speeding up computations, and when employed appropriately it can lead to little loss in classification quality.
Style transfer is a process of transferring a style from one image to an other image, obtaining a new image which is an artistic mixture of the two. Recent work on this problem involved adopting convolutional neural-networks (CNN). This ignited a renewed interest in this field, due to the very impressive results obtained. Gaty’s approach utilizes this method. There exists an alternative path toward handling the style transfer task via generalization of texture synthesis algorithms. Kwatra’s approach utilizes this method. We shall implement style transfer using Kwatra’s and Gaty’s approach and compare their performance. Based on the results obtained, after using 2 metrics, we shall find out which works the best.
A Gröbner basis algorithm computes a good basis for an ideal of a polynomial ring and appears in various situations of cryptography. In particular, it has been used in the security analysis of multivariate public key cryptography (MPKC), and has been studied for a long time; however, it is far from a complete understanding. We consider the algebraic attack using a Gröbner basis algorithm for a new multivariate encryption scheme proposed by Jiahui Chen et al. at Theoretical Computer Science 2020. Their idea to construct a new scheme was to use the minus and plus modifiers to prevent known attacks, such as linearization attacks. Moreover, they discussed having a resistance to the algebraic attack using a Gröbner basis algorithm. However, in our experiments, the algebraic attack breaks their claimed 80- and 128-bit security parameters in reasonable times. It is necessary to understand whether their scheme can avoid such an attack by introducing a slight modification. In this paper, we theoretically describe why the algebraic attack breaks their scheme and give a precise complexity of the algebraic attack. As a result, we demonstrate that the algebraic attack can break the claimed 80- and 128-bit security parameters in the complexities of approximately 25 and 32 bits, respectively. Moreover, based on our complexity estimation of the algebraic attack, we conclude that the Chen et al. scheme is not practical.
For an odd prime p, let E_0 be a supersingular elliptic curve over 𝔽_p^2 with . The Deuring correspondence gives a one-to-one correspondence between isogenies E_0 ⟶ E and left -ideals. In 2014, Kohel–Lauter–Petit–Tignol provided a probabilistic algorithm, called the KLPT algorithm, that finds an equivalent ideal of a given left -ideal with some powersmooth norm. It is useful for both cryptanalyses and constructions in supersingular isogeny-based cryptography. In this paper, we modify the original KLPT algorithm to improve its output quality so that an output ideal has smaller norm. This would give an efficiency for the constructive Deuring correspondence, in which we compute the supersingular elliptic curve corresponding to a given left -ideal via the Deuring correspondence. We also report implementation results of our modified KLPT algorithm for primes p up to around 45 bits. This is the largest scale of implementation reports for the original KLPT algorithm in the literature.
In this paper, we discuss the boundedness of generalized Libera operator ^γ on mixed-norm spaces H^p,q_α ,ν . As a consequence, we find few results about the action of the operator ^γ on various function spaces such as Hardy, Zygmund, Lipschitz, Bloch type, and Besov spaces.
In this work, Improved Meadow Fritillary Butterfly (IMB) optimization algorithm is designed for power loss reduction. In the Meadow Fritillary Butterfly optimization algorithm, the exploration method has two properties of Meadow Fritillary butterflies; in that butterfly adjusting operator in preliminary iterations remarkably directs the exploration procedure in the direction of the present most outstanding solution. To trounce this deficit, firefly’s algorithm exploration method has been incorporated into the standard Meadow Fritillary optimization algorithm. Meadow Fritillary Butterfly (IMB) optimization algorithm validated in IEEE 30 and 57 bus test systems and Decline in loss has been attained.
Time synchronization in underwater sensor networks (UWS-Ns) has come up as an active area of research in recent years. The drastic differences in physical and chemical characteristics of underwater environments as compared to terrestrial environments have necessitated that underwater time synchronization protocols be different from their terrestrial counterparts. The protocols reviewed in this survey range from static to dynamic, doppler assisted, and ordinary least squares-based protocols. This paper is an attempt to review ten such UWSN time synchronization protocols that are in extant literature and elaborate on the advantages and disadvantages of each of them.
Authentication of public keys in asymmetric cryptography is an important objective to fulfil. Identity-based schemes are an efficient solution for this scenario. There are various identity-based schemes that are based on either factoring problem or DLP problem, but these problems can be solved in polynomial time using quantum computers. Multivariate cryptography is one of the main alternatives for the construction of post-quantum digital signature schemes. Digital signature construction needs modest computations in multivariate public key cryptography, so these schemes are considered quite efficient. In this paper, we present a new identity-based signature scheme in which the signature size and user key size are relatively small. The design of our proposed signature scheme is based on MQDSS construction.
The $$L(k_1,k_2)$$ labeling is a type of labeling of vertices of a graph G(V, E) where absolute difference between labels of vertices at graph distances 1 and 2 are at least $$k_1$$ and $$k_2$$ , respectively, where $$k_1$$ and $$k_2$$ are two non-negative real numbers. The objective of $$L(k_1,k_2)$$ labeling problem is to find a labeling f of the vertices of G such that the span is minimized, where span $$\lambda (G;f)$$ of a labeling f is defined as $$\displaystyle \max _{u \in V} f(u) - \min _{v \in V} f(v)$$ . The motivation for $$L(k_1,k_2)$$ labeling problem for infinite triangular grid $$T_\varDelta $$ comes from the fact that a class of frequency assignment problems in cellular networks can be modeled as $$L(k_1,k_2)$$ labeling problems of $$T_\varDelta $$ . Existing bounds on $$\lambda (T_\varDelta ;f)$$ for $$k_1 \le k_2$$ are based on enumerations through partial computer simulations. In this paper, we attempt to derive $$\lambda (T_\varDelta ;f)$$ theoretically by exploring the underlined graph structures without any computer simulations. Specifically, we first select a sub graph of $$T_\varDelta $$ which holds certain structural properties and then using these properties, we find the lower bound of span of the concerned subgraph. This finally leads us to obtain the lower bound of $$\lambda (T_\varDelta ;f)$$ . We establish that $$\lambda (T_\varDelta ;f) \ge 3+2h$$ when $$h < 1/2$$ and $$\lambda (T_\varDelta ;f) \ge 4$$ when $$h \ge 1/2 $$ , where $$h=k_1/k_2$$ . Ours is the first attempt to derive $$\lambda (T_\varDelta ;f)$$ theoretically and our obtained results exactly coincide with that of the known bound obtained through computer simulations when $$0 \le h \le 1/3$$ . For $$h > 1/3$$ , such known bounds are actually finer than ours.
