Search

Rising breast cancer (BC) occurrence and mortality are major global concerns for women. Deep learning (DL) has demonstrated superior diagnostic performance in BC classification compared to human expert readers. However, the predominant use of unimodal (digital mammography) features may limit the current performance of diagnostic models. To address this, we collected a novel multimodal dataset comprising both imaging and textual data. This study proposes a multimodal DL architecture for BC classification, utilising images (mammograms; four views) and textual data (radiological reports) from our new in-house dataset. Various augmentation techniques were applied to enhance the training data size for both imaging and textual data. We explored the performance of eleven SOTA DL architectures (VGG16, VGG19, ResNet34, ResNet50, MobileNet-v3, EffNet-b0, EffNet-b1, EffNet-b2, EffNet-b3, EffNet-b7, and Vision Transformer (ViT)) as imaging feature extractors. For textual feature extraction, we utilised either artificial neural networks (ANNs) or long short-term memory (LSTM) networks. The combined imaging and textual features were then inputted into an ANN classifier for BC classification, using the late fusion technique. We evaluated different feature extractor and classifier arrangements. The VGG19 and ANN combinations achieved the highest accuracy of 0.951. For precision, the VGG19 and ANN combination again surpassed other CNN and LSTM, ANN based architectures by achieving a score of 0.95. The best sensitivity score of 0.903 was achieved by the VGG16+LSTM. The highest F1 score of 0.931 was achieved by VGG19+LSTM. Only the VGG16+LSTM achieved the best area under the curve (AUC) of 0.937, with VGG16+LSTM closely following with a 0.929 AUC score., Comment: The paper was accepted and presented in 3rd Workshop on Cancer Prevention, detection, and intervenTion (CaPTion @ MICCAI 2024)

Steganography in multimedia aims to embed secret data into an innocent looking multimedia cover object. This embedding introduces some distortion to the cover object and produces a corresponding stego object. The embedding distortion is measured by a cost function that determines the detection probability of the existence of the embedded secret data. A cost function related to the maximum embedding rate is typically employed to evaluate a steganographic system. In addition, the distribution of multimedia sources follows the Gibbs distribution which is a complex statistical model that restricts analysis. Thus, previous multimedia steganographic approaches either assume a relaxed distribution or presume a proposition on the maximum embedding rate and then try to prove it is correct. Conversely, this paper introduces an analytic approach to determining the maximum embedding rate in multimedia cover objects through a constrained optimization problem concerning the relationship between the maximum embedding rate and the probability of detection by any steganographic detector. The KL-divergence between the distributions for the cover and stego objects is used as the cost function as it upper bounds the performance of the optimal steganographic detector. An equivalence between the Gibbs and correlated-multivariate-quantized-Gaussian distributions is established to solve this optimization problem. The solution provides an analytic form for the maximum embedding rate in terms of the WrightOmega function. Moreover, it is proven that the maximum embedding rate is in agreement with the commonly used Square Root Law (SRL) for steganography, but the solution presented here is more accurate. Finally, the theoretical results obtained are verified experimentally., Comment: arXiv admin note: substantial text overlap with arXiv:2111.04960

The advent of Bitcoin, and consequently Blockchain, has ushered in a new era of decentralization. Blockchain enables mutually distrusting entities to work collaboratively to attain a common objective. However, current Blockchain technologies lack scalability, which limits their use in Internet of Things (IoT) applications. Many devices on the Internet have the computational and communication capabilities to facilitate decision-making. These devices will soon be a 50 billion node network. Furthermore, new IoT business models such as Sensor-as-a-Service (SaaS) require a robust Trust and Reputation System (TRS). In this paper, we introduce an innovative distributed ledger combining Tangle and Blockchain as a TRS framework for IoT. The combination of Tangle and Blockchain provides maintainability of the former and scalability of the latter. The proposed ledger can handle large numbers of IoT device transactions and facilitates low power nodes joining and contributing. Employing a distributed ledger mitigates many threats, such as whitewashing attacks. Along with combining payments and rating protocols, the proposed approach provides cleaner data to the upper layer reputation algorithm.

Dealing with imbalanced data is one of the main challenges in machine/deep learning algorithms for classification. This issue is more important with log message data as it is typically very imbalanced and negative logs are rare. In this paper, a model is proposed to generate text log messages using a SeqGAN network. Then features are extracted using an Autoencoder and anomaly detection is done using a GRU network. The proposed model is evaluated with two imbalanced log data sets, namely BGL and Openstack. Results are presented which show that oversampling and balancing data increases the accuracy of anomaly detection and classification., Comment: 14 pages, 4 figures, 2 tables

Log messages are now widely used in software systems. They are important for classification as millions of logs are generated each day. Most logs are unstructured which makes classification a challenge. In this paper, Deep Learning (DL) methods called Auto-LSTM, Auto-BLSTM and Auto-GRU are developed for anomaly detection and log classification. These models are used to convert unstructured log data to extracted features which is suitable for classification algorithms. They are evaluated using four data sets, namely BGL, Openstack, Thunderbird and IMDB. The first three are popular log data sets while the fourth is a movie review data set which is used for sentiment classification. The results obtained show that Auto-LSTM, Auto-BLSTM and Auto-GRU perform better than other well-known algorithms., Comment: 18 pages, 5 figures, 3 tables

This paper considers the construction of isodual quasi-cyclic codes. First we prove that two quasi-cyclic codes are permutation equivalent if and only if their constituent codes are equivalent. This gives conditions on the existence of isodual quasi-cyclic codes. Then these conditions are used to obtain isodual quasi-cyclic codes. We also provide a construction for isodual quasi-cyclic codes as the matrix product of isodual codes., Comment: arXiv admin note: substantial text overlap with arXiv:1607.07798

In this paper, we give conditions for the existence of Hermitian self-dual $\Theta-$cyclic and $\Theta-$negacyclic codes over the finite chain ring $\mathbb{F}_q+u\mathbb{F}_q$. By defining a Gray map from $R=\mathbb{F}_q+u\mathbb{F}_q$ to $\mathbb{F}_{q}^{2}$, we prove that the Gray images of skew cyclic codes of odd length $n$ over $R$ with even characteristic are equivalent to skew quasi-twisted codes of length $2n$ over $\mathbb{F}_q$ of index $2$. We also extend an algorithm of Boucher and Ulmer \cite{BF3} to construct self-dual skew cyclic codes based on the least common left multiples of non-commutative polynomials over $\mathbb{F}_q+u\mathbb{F}_q$.

In this paper, a linear $\ell$-intersection pair of codes is introduced as a generalization of linear complementary pairs of codes. Two linear codes are said to be a linear $\ell$-intersection pair if their intersection has dimension $\ell$. Characterizations and constructions of such pairs of codes are given in terms of the corresponding generator and parity-check matrices. Linear $\ell$-intersection pairs of MDS codes over $\mathbb{F}_q$ of length up to $q+1$ are given for all possible parameters. As an application, linear $\ell$-intersection pairs of codes are used to construct entanglement-assisted quantum error correcting codes. This provides a large number of new MDS entanglement-assisted quantum error correcting codes.

Stack Overflow is the most popular Q&A website among software developers. As a platform for knowledge sharing and acquisition, the questions posted in Stack Overflow usually contain a code snippet. Stack Overflow relies on users to properly tag the programming language of a question and it simply assumes that the programming language of the snippets inside a question is the same as the tag of the question itself. In this paper, we propose a classifier to predict the programming language of questions posted in Stack Overflow using Natural Language Processing (NLP) and Machine Learning (ML). The classifier achieves an accuracy of 91.1% in predicting the 24 most popular programming languages by combining features from the title, body and the code snippets of the question. We also propose a classifier that only uses the title and body of the question and has an accuracy of 81.1%. Finally, we propose a classifier of code snippets only that achieves an accuracy of 77.7%. These results show that deploying Machine Learning techniques on the combination of text and the code snippets of a question provides the best performance. These results demonstrate also that it is possible to identify the programming language of a snippet of few lines of source code. We visualize the feature space of two programming languages Java and SQL in order to identify some special properties of information inside the questions in Stack Overflow corresponding to these languages.

Determining the programming language of a source code file has been considered in the research community; it has been shown that Machine Learning (ML) and Natural Language Processing (NLP) algorithms can be effective in identifying the programming language of source code files. However, determining the programming language of a code snippet or a few lines of source code is still a challenging task. Online forums such as Stack Overflow and code repositories such as GitHub contain a large number of code snippets. In this paper, we describe Source Code Classification (SCC), a classifier that can identify the programming language of code snippets written in 21 different programming languages. A Multinomial Naive Bayes (MNB) classifier is employed which is trained using Stack Overflow posts. It is shown to achieve an accuracy of 75% which is higher than that with Programming Languages Identification (PLI a proprietary online classifier of snippets) whose accuracy is only 55.5%. The average score for precision, recall and the F1 score with the proposed tool are 0.76, 0.75 and 0.75, respectively. In addition, it can distinguish between code snippets from a family of programming languages such as C, C++ and C#, and can also identify the programming language version such as C# 3.0, C# 4.0 and C# 5.0.

In this paper we investigate linear codes with complementary dual (LCD) codes and formally self-dual codes over the ring $R=\F_{q}+v\F_{q}+v^{2}\F_{q}$, where $v^{3}=v$, for $q$ odd. We give conditions on the existence of LCD codes and present construction of formally self-dual codes over $R$. Further, we give bounds on the minimum distance of LCD codes over $\F_q$ and extend these to codes over $R$., Comment: 19 pages

We study the performance of ternary isodual codes which are not self-dual and ternary self-dual codes, as measured by the decoding error probability in bounded distance decoding. We compare the performance of ternary double circulant and double twistulant codes which are not self-dual with ternary extremal self-dual codes. We also investigate the performance of ternary self-dual codes having large minimum weights., Comment: 19 pages. arXiv admin note: text overlap with arXiv:1604.06878

This paper considers the equivalence problem for quasi-cyclic codes over finite fields. The results obtained are used to construct isodual quasi-cyclic codes.

Entanglement-assisted quantum error correcting codes (EAQECCs) are a simple and fundamental class of codes. They allow for the construction of quantum codes from classical codes by relaxing the duality condition and using pre-shared entanglement between the sender and receiver. However, in general it is not easy to determine the number of shared pairs required to construct an EAQECC. In this paper, we show that this number is related to the hull of the classical code. Using this fact, we give methods to construct EAQECCs requiring desirable amount of entanglement. This leads to design families of EAQECCs with good error performance. Moreover, we construct maximal entanglement EAQECCs from LCD codes. Finally, we prove the existence of asymptotically good EAQECCs in the odd characteristic case.

In this note, we investigate the performance of optimal double circulant even codes which are not self-dual, as measured by the decoding error probability in bounded distance decoding. To do this, we classify the optimal double circulant even codes that are not self-dual which have the smallest weight distribution for lengths up to $72$. We also give some restrictions on the weight enumerators of (extremal) self-dual $[54,27,10]$ codes with shadows of minimum weight $3$. Finally, we consider the performance of extremal self-dual codes of lengths $88$ and $112$., Comment: 14 pages

A classification of extremal double circulant self-dual codes of lengths up to $88$ is known. We give a classification of extremal double circulant self-dual codes of lengths $90,92,94$ and $96$. We also classify double circulant self-dual codes with parameters $[90,45,14]$ and $[96,48,16]$. In addition, we demonstrate that no double circulant self-dual $[90,45,14]$ code has an extremal self-dual neighbor, and no double circulant self-dual $[96,45,16]$ code has a self-dual neighbor with minimum weight at least $18$., Comment: 15 pages

In this paper, we construct quantum synchronizable codes (QSCs) based on the sum and intersection of cyclic codes. Further, infinite families of QSCs are obtained from BCH and duadic codes. Moreover, we show that the work of Fujiwara~\cite{fujiwara1} can be generalized to repeated root cyclic codes (RRCCs) such that QSCs are always obtained, which is not the case with simple root cyclic codes. The usefulness of this extension is illustrated via examples of infinite families of QSCs from repeated root duadic codes. Finally, QSCs are constructed from the product of cyclic codes.

In this paper, we construct linear codes over $\mathbb{Z}_4$ with bounded $GC$-content. The codes are obtained using a greedy algorithm over $\mathbb{Z}_4$. Further, upper and lower bounds are derived for the maximum size of DNA codes of length $n$ with constant $GC$-content $w$ and edit distance $d$.

In this paper, we introduce the homogeneous weight and homogeneous Gray map over the ring $R_{q}=\mathbb{F}_{2}[u_{1},u_{2},\ldots,u_{q}]/\left\langle u_{i}^{2}=0,u_{i}u_{j}=u_{j}u_{i}\right\rangle$ for $q \geq 2$. We also consider the construction of simplex and MacDonald codes of types $\alpha$ and $\beta$ over this ring.

In this paper, we give an important isomorphism between contacyclic codes and cyclic codes over finite principal ideal rings. Necessary and sufficient conditions for the existence of non-trivial cyclic self-dual codes over finite principal ideal rings are given.

In this paper we investigate the structure of repeated root constacyclic codes of length $2^amp^r$ over $\mathbb{F}_{p^s}$ with $a\geq1$ and $(m,p)=1$. We characterize the codes in terms of their generator polynomials. This provides simple conditions on the existence of self-dual negacyclic codes. Further, we gave cases where the constacyclic codes are equivalent to cyclic codes., Comment: arXiv admin note: text overlap with arXiv:1406.1848 by other authors

In this paper we define $\mathbb{Z}_{2}\mathbb{Z}_{4}-$Simplex and MacDonald Codes of type $\alpha $ and $\beta $ and we give the covering radius of these codes., Comment: arXiv admin note: text overlap with arXiv:1411.1822 by other authors

In this paper we extend the work of Lisonek and Singh on construction X for quantum error-correcting codes to finite fields of order $p^2^ where p is prime. The results obtained are applied to the dual of Hermitian repeated root cyclic codes to generate new quantum error-correcting codes.

An algorithm is presented for approximating a single-user channel with a prime input alphabet size. The result is an upgraded version of the channel with a reduced output alphabet size. It is shown that this algorithm can be used to reduce the output alphabet size to the input alphabet size in most cases.

In a distributed storage systems (DSS), regenerating codes are used to optimize bandwidth in the repair process of a failed node. To optimize other DSS parameters such as computation and disk I/O, Distributed Replication-based Simple Storage (Dress) Codes consisting of an inner Fractional Repetition (FR) code and an outer MDS code are commonly used. Thus constructing FR codes is an important research problem, and several constructions using graphs and designs have been proposed. In this paper, we present an algorithm for constructing the node-packet distribution matrix of FR codes and thus enumerate some FR codes up to a given number of nodes n. We also present algorithms for constructing regular graphs which give rise to FR codes., Comment: 6 pages, 4 figures, Submitted to IEEE Netcod 2013, Code List at http://www.ece.uvic.ca/~agullive/manish/List.html

This paper present the construction cyclic isodual codes over finite fields and finite chain rings. These codes are monomially equivalent to their dual. Conditions are given for the existence of cyclic isodual codes. In addition, the concept of duadic codes over finite fields is extended to finite chain rings. Several constructions of isodual cyclic codes and self-dual codes are given.

The Radio Environment Map (REM) provides an effective approach to Dynamic Spectrum Access (DSA) in Cognitive Radio Networks (CRNs). Previous results on REM construction show that there exists a tradeoff between the number of measurements (sensors) and REM accuracy. In this paper, we analyze this tradeoff and determine that the REM error is a decreasing and convex function of the number of measurements (sensors). The concept of geographic entropy is introduced to quantify this relationship. And the influence of sensor deployment on REM accuracy is examined using information theory techniques. The results obtained in this paper are applicable not only for the REM, but also for wireless sensor network deployment., Comment: 6 pages, 7 figures, IEEE WCNC conference

The definition of polar codes given by Arikan is explicit, but the construction complexity is an issue. This is due to the exponential growth in the size of the output alphabet of the bit-channels as the codeword length increases. Tal and Vardy recently presented a method for constructing polar codes which controls this growth. They approximated each bit-channel with a better channel and a worse channel while reducing the alphabet size. They constructed a polar code based on the worse channel and used the better channel to measure the distance from the optimal channel. This paper considers the knowledge gained from the perspective of the better channel. A method is presented using iterative upgrading of the bit-channels which successively results in a channel closer to the original one. It is shown that this approach can be used to obtain a channel arbitrarily close to the original channel, and therefore to the optimal construction of a polar code., Comment: arXiv admin note: text overlap with arXiv:1105.6164 by other authors

For $\lambda$ an $n$-th power of a unit in a finite chain ring we prove that $\lambda$-constacyclic repeated-root codes over some finite chain rings are equivalent to cyclic codes. This allows us to simplify the structure of some constacylic codes. We also study the $\alpha +p \beta$-constacyclic codes of length $p^s$ over the Galois ring $GR(p^e,r)$.

We give the structure of $\lambda$-constacyclic codes of length $p^sm$ over $R=\mathbb{F}_{p^r}+u \mathbb{F}_{p^r}+...+ u^{e-1}\mathbb{F}_{p^r}$ with $\lambda \in \F_{p^r}^*$. We also give the structure of $\lambda$-constacyclic codes of length $p^sm$ with $\lambda=\alpha_1+u\alpha_2+...+u^{e-1} \alpha_{e-1}$, where $\alpha_1,\alpha_2 \neq 0$ and study the self-duality of these codes.

This paper considers cyclic DNA codes of arbitrary length over the ring $R=\F_2[u]/u^4-1$. A mapping is given between the elements of $R$ and the alphabet $\{A,C,G,T\}$ which allows the additive stem distance to be extended to this ring. Cyclic codes over $R$ are designed such that their images under the mapping are also cyclic or quasi-cyclic of index 2. The additive distance and hybridization energy are functions of the neighborhood energy., Comment: there is an error in Lemma 3.4

In this paper, we consider the construction of linear lexicodes over finite chain rings by using a $B$-ordering over these rings and a selection criterion. % and a greedy Algorithm. As examples we give lexicodes over $\mathbb{Z}_4$ and $\mathbb{F}_2+u\mathbb{F}_2$. %First, greedy algorithms are presented to construct %lexicodes using a multiplicative property. Then, greedy algorithms %are given for the case when the selection criteria is not %multiplicative such as the minimum distance constraint. It is shown that this construction produces many optimal codes over rings and also good binary codes. Some of these codes meet the Gilbert bound. We also obtain optimal self-dual codes, in particular the octacode.

We construct codes over the ring $\mathbb{F}_2+u\mathbb{F}_2$ with $u^2=0$. These code are designed for use in DNA computing applications. The codes obtained satisfy the reverse complement constraint, the $GC$ content constraint and avoid the secondary structure. they are derived from the cyclic complement reversible codes over the ring $\mathbb{F}_2+u\mathbb{F}_2$. We also construct an infinite family of BCH DNA codes.

In this paper we investigate repeated root cyclic and negacyclic codes of length $p^rm$ over $\mathbb{F}_{p^s}$ with $(m,p)=1$. In the case $p$ odd, we give necessary and sufficient conditions on the existence of negacyclic self-dual codes. When $m=2m'$ with $m'$ odd, we characterize the codes in terms of their generator polynomials. This provides simple conditions on the existence of self-dual negacyclic codes, and generalizes the results of Dinh \cite{dinh}. We also answer an open problem concerning the number of self-dual cyclic codes given by Jia et al. \cite{jia}.

In this paper we give the structure of constacyclic codes over formal power series and chain rings. We also present necessary and sufficient conditions on the existence of MDS codes over principal ideal rings. These results allow for the construction of infinite families of MDS self-dual codes over finite chain rings, formal power series and principal ideal rings.

We determine the permutation groups that arise as the automorphism groups of cyclic combinatorial objects. As special cases we classify the automorphism groups of cyclic codes. We also give the permutations by which two cyclic combinatorial objects on $p^m$ elements are equivalent.

New infinite families of quantum symmetric and asymmetric codes are constructed. Several of these are MDS. The codes obtained are shown to have parameters which are better than previously known. A number of known codes are special cases of the codes given here.

This paper considers the use an entanglement breaking channel in the construction of a secure bit-commitment protocol. It is shown that this can be done via a depolarizing quantum channel.

In this paper, we introduce a new quantum bit commitment protocol which is practically secure against entanglement attacks. A general cheating strategy is discussed and shown to be practically ineffective against the proposed approach.