On Turing Machines Deciding According to the Shortest Computations

In this paper we propose and analyse from the computational complexity point of view several new variants of nondeterministic Turing machines. In the first such variant, a machine accepts a given input word if and only if one of its shortest possible computations on that word is accepting; on the other hand, the machine rejects the input word when all the shortest computations performed by the machine on that word are rejecting. We are able to show that the class of languages decided in polynomial time by such machines is PNP[log]. When we consider machines that decide a word according to the decision taken by the lexicographically first shortest computation, we obtain a new characterization of PNP. A series of other ways of deciding a language with respect to the shortest computations of a Turing machine are also discussed.

Complete Abstractions for Checking Language Inclusion

We study the language inclusion problem L 1 ⊆ L 2 , where L 1 is regular or context-free. Our approach relies on abstract interpretation and checks whether an overapproximating abstraction of L 1 , obtained by approximating the Kleene iterates of its least fixpoint characterization, is included in L 2 . We show that a language inclusion problem is decidable whenever this overapproximating abstraction satisfies a completeness condition (i.e., its loss of precision causes no false alarm) and prevents infinite ascending chains (i.e., it guarantees termination of least fixpoint computations). This overapproximating abstraction of languages can be defined using quasiorder relations on words, where the abstraction gives the language of all the words “greater than or equal to” a given input word for that quasiorder. We put forward a range of such quasiorders that allow us to systematically design decision procedures for different language inclusion problems, such as regular languages into regular languages or into trace sets of one-counter nets, and context-free languages into regular languages. In the case of inclusion between regular languages, some of the induced inclusion checking procedures correspond to well-known state-of-the-art algorithms, like the so-called antichain algorithms. Finally, we provide an equivalent language inclusion checking algorithm based on a greatest fixpoint computation that relies on quotients of languages and, to the best of our knowledge, was not previously known.

Space Complexity of Stack Automata Models

This paper examines several measures of space complexity of variants of stack automata: non-erasing stack automata and checking stack automata. These measures capture the minimum stack size required to accept every word in the language of the automaton (weak measure), the maximum stack size used in any accepting computation on any accepted word (accept measure), and the maximum stack size used in any computation (strong measure). We give a detailed characterization of the accept and strong space complexity measures for checking stack automata. Exactly one of three cases can occur: the complexity is either bounded by a constant, behaves like a linear function, or it can not be bounded by any function of the length of the input word (and it is decidable which case occurs). However, this result does not hold for non-erasing stack automata; we provide an example where the space complexity grows proportionally to the square root of the length of the input. Furthermore, we study the complexity bounds of machines which accept a given language, and decidability of space complexity properties.

AHWR-Net: offline handwritten amharic word recognition using convolutional recurrent neural network

Amharic ("Image missing") is the official language of the Federal Government of Ethiopia, with more than 27 million speakers. It uses an Ethiopic script, which has 238 core and 27 labialized characters. It is a low-resourced language, and a few attempts have been made so far for its handwritten text recognition. However, Amharic handwritten text recognition is challenging due to the very high similarity between characters. This paper presents a convolutional recurrent neural networks based offline handwritten Amharic word recognition system. The proposed framework comprises convolutional neural networks (CNNs) for feature extraction from input word images, recurrent neural network (RNNs) for sequence encoding, and connectionist temporal classification as a loss function. We designed a custom CNN model and compared its performance with three different state-of-the-art CNN models, including DenseNet-121, ResNet-50 and VGG-19 after modifying their architectures to fit our problem domain, for robust feature extraction from handwritten Amharic word images. We have conducted detailed experiments with different CNN and RNN architectures, input word image sizes, and applied data augmentation techniques to enhance performance of the proposed models. We have prepared a handwritten Amharic word dataset, HARD-I, which is available publicly for researchers. From the experiments on various recognition models using our dataset, a WER of 5.24 % and CER of 1.15 % were achieved using our best-performing recognition model. The proposed models achieve a competitive performance compared to existing models for offline handwritten Amharic word recognition.

Performance analysis of Low energy and highspeed DA-RNS based FIR filter design for SDR Applications on FPGA

For different applications, the Finite Impulse Response (FIR) filter is widely used in digital signal processing (DSP) applications. We exhibit a significant Residue Number System (RNS)-based FIR filter design for Software Defined Radio (SDR) filtration in this article. Including its underlying concurrency and information clustering process, the RNS provides important statistics over FIR application in specific. According to several residue computing and reverse translation, expanded bit size results in a significant performance trade-off, conversely. Through RNS replication, accompanied by conditional delay optimized reverse processing to minimize the FIR filter trade-off features with filter duration optimized Residue Number System arithmetic is proposed in this study, which involves distributed arithmetic-based residue processing. To execute the task of reverse translation and to store pre-computational properties, the suggested Residue Number System architecture makes use of built-in RAM blocks found in field-programmable gate array (FPGA) devices. The proposed FIR filter with core optimized RNS has the benefit of lowering processing latency delay while rising performance torque. Followed by FPGA hardware synthesis for different input word sizes and FIR lengths verification by the efficiency of the FIR filter core, fetal audio signal detection is performed first. The test results reveal that over the optimization procedure RNS method, a compromise in traditional RNS FIR over filter size is narrowed, as well as a substantial decrease in sophistication.

FPGA Implementation of high speed-low energy RNS based Reconfigurable-FIR Filter for Cognitive Radio Applications

The Finite impulse response (FIR) filter is prominently employed in many digital signal processing (DSP) systems for various applications. In this paper, we present a high-performance RNS based FIR filter design for filtration in SDR applications. In general, the residue number system (RNS) gives significant metrics over FIR implementation with its inherent parallelism and data partitioning mechanism. But with increased bit width cause considerable performance trade-off due to both residue computation and reverse conversion. In this paper optimized Residue Number System (RNS) arithmetic is proposed which includes distributed arithmetic based residue computation during RNS multiplication followed by speculative delay optimized reverse computation to mitigate the FIR filter trade-off characteristics with filter length. The proposed RNS design utilizes built-in RAMs block present in the devices of FPGA to accomplish the process of reverse conversion and to store pre-computational values. A distinctive feature of the proposed FIR filter implementation with core optimized RNS is to minimize hardware complexity overhead with the improved operating speed. Initially, fetal audio signal detection is carried out to validate the functionality of FIR filter core and FPGA hardware synthesis is carried out for various input word size and FIR length. From the experimental, it is proved that the trade-off exists in conventional RNS FIR over filter length is narrow down along with considerable complexity reduction with our proposed optimized RNS system.

Languages Accepted by Weighted Restarting Automata*

Weighted restarting automata have been introduced to study quantitative aspects of computations of restarting automata. In earlier works we studied the classes of functions and relations that are computed by weighted restarting automata. Here we use them to define classes of formal languages by restricting the weight associated to a given input word through an additional requirement. In this way, weighted restarting automata can be used as language acceptors. First, we show that by using the notion of acceptance relative to the tropical semiring, we can avoid the use of auxiliary symbols. Furthermore, a certain type of word-weighted restarting automata turns out to be equivalent to non-forgetting restarting automata, and another class of languages accepted by word-weighted restarting automata is shown to be closed under the operation of intersection. This is the first result that shows that a class of languages defined in terms of a quite general class of restarting automata is closed under intersection. Finally, we prove that the restarting automata that are allowed to use auxiliary symbols in a rewrite step, and to keep on reading after performing a rewrite step can be simulated by regular-weighted restarting automata that cannot do this.

Self-Verifying Pushdown and Queue Automata

We study the computational and descriptional complexity of self-verifying pushdown automata (SVPDA) and self-verifying realtime queue automata (SVRQA). A self-verifying automaton is a nondeterministic device whose nondeterminism is symmetric in the following sense. Each computation path can give one of the answers yes, no, or do not know. For every input word, at least one computation path must give either the answer yes or no, and the answers given must not be contradictory. We show that SVPDA and SVRQA are automata characterizations of so-called complementation kernels, that is, context-free or realtime nondeterministic queue automaton languages whose complement is also context free or accepted by a realtime nondeterministic queue automaton. So, the families of languages accepted by SVPDA and SVRQA are strictly between the families of deterministic and nondeterministic languages. Closure properties and various decidability problems are considered. For example, it is shown that it is not semidecidable whether a given SVPDA or SVRQA can be made self-verifying. Moreover, we study descriptional complexity aspects of these machines. It turns out that the size trade-offs between nondeterministic and self-verifying as well as between self-verifying and deterministic automata are non-recursive. That is, one can choose an arbitrarily large recursive function f, but the gain in economy of description eventually exceeds f when changing from the former system to the latter.

Máquina de Turing Analógica para Ensino de Linguagens Formais e Autômatos

The article describes the development of a practical device for teachingin the area of Computer Theory. In the study, an adaptationof the Turing Machine is presented, using hardware and softwareintegration to interpret Formal Languages. Simulating an Automaton,sensors and motors are used to move the device head to the leftand right and to read and write the input tape. The developmentof the mechanism is described in two parts, the first includes thehardware that consists of the construction and adaptation of theTuring Machine, the second the implementation of the software andcommunication part between both. The developed device, allowsthe interpretation of a binary alphabet (0, 1), where an input word isaccepted, and as an output result, such device rejected or acceptedthe word.

Development of Integrated Neural Network Model for Identification of Fake Reviews in E-Commerce Using Multidomain Datasets

Online product reviews play a major role in the success or failure of an E-commerce business. Before procuring products or services, the shoppers usually go through the online reviews posted by previous customers to get recommendations of the details of products and make purchasing decisions. Nevertheless, it is possible to enhance or hamper specific E-business products by posting fake reviews, which can be written by persons called fraudsters. These reviews can cause financial loss to E-commerce businesses and misguide consumers to take the wrong decision to search for alternative products. Thus, developing a fake review detection system is ultimately required for E-commerce business. The proposed methodology has used four standard fake review datasets of multidomains include hotels, restaurants, Yelp, and Amazon. Further, preprocessing methods such as stopword removal, punctuation removal, and tokenization have performed as well as padding sequence method for making the input sequence has fixed length during training, validation, and testing the model. As this methodology uses different sizes of datasets, various input word-embedding matrices of n-gram features of the review’s text are developed and created with help of word-embedding layer that is one component of the proposed model. Convolutional and max-pooling layers of the CNN technique are implemented for dimensionality reduction and feature extraction, respectively. Based on gate mechanisms, the LSTM layer is combined with the CNN technique for learning and handling the contextual information of n-gram features of the review’s text. Finally, a sigmoid activation function as the last layer of the proposed model receives the input sequences from the previous layer and performs binary classification task of review text into fake or truthful. In this paper, the proposed CNN-LSTM model was evaluated in two types of experiments, in-domain and cross-domain experiments. For an in-domain experiment, the model is applied on each dataset individually, while in the case of a cross-domain experiment, all datasets are gathered and put into a single data frame and evaluated entirely. The testing results of the model in-domain experiment datasets were 77%, 85%, 86%, and 87% in the terms of accuracy for restaurant, hotel, Yelp, and Amazon datasets, respectively. Concerning the cross-domain experiment, the proposed model has attained 89% accuracy. Furthermore, comparative analysis of the results of in-domain experiments with existing approaches has been done based on accuracy metric and, it is observed that the proposed model outperformed the compared methods.