In recent years, there has been a meteoric rise in the use of small computing devices, including RFID (Radio Frequency Identification) tags, wireless sensors, embedded electronics, and IoT (Internet of Things) gadgets...
In recent years, there has been a meteoric rise in the use of small computing devices, including RFID (Radio Frequency Identification) tags, wireless sensors, embedded electronics, and IoT (Internet of Things) gadgets. Given that these gadgets handle highly sensitive data, their protection should be a top priority. Therefore, it is crucial to implement a secure encryption system on susceptible devices. Standard encryption algorithms like RSA and AES are resource-intensive, require substantial memory, and can slow down devices. In this work, we introduce a lightweight cryptosystem that is suitable for low-powered devices without compromising security. This system utilizes Substitution-Permutation network, incorporating a variation of the Feistel design, to establish a secure symmetric key block cipher. As an enhancement over the standard technique, the proposed algorithm requires less run-time while still providing sufficient protection to maintain an avalanche effect of nearly 50%.
The goal of this research is to aid doctors in the diagnosis of PCOS in female patients. Diagnosing the condition in question depends on many factors making it complex to diagnose. The model developed would help confi...
The goal of this research is to aid doctors in the diagnosis of PCOS in female patients. Diagnosing the condition in question depends on many factors making it complex to diagnose. The model developed would help confirm a doctor's diagnosis to further its reliability. The model tested several classifiers, including Extreme Gradient Boosting (XGBoost), Linear Discriminant Analysis (LDA), and Adaptive Boosting (Ada-Boost). The highest accuracy was 94.4% using the Random Forest classifier with the Bagging method. This accuracy surpasses any previously achieved results using the same dataset, which were 91% and 92%. The results achieved were using a 10-Fold cross-validation.
Many organizations use tendering including government to obtain goods and services from the providers of the service and manufacturing companies. There has been an advance evolution in the process from traditional har...
Many organizations use tendering including government to obtain goods and services from the providers of the service and manufacturing companies. There has been an advance evolution in the process from traditional hardcopy based to modern electronic based tendering. Since in the present scenario internet is being used there are many security implications that can be associated with it. Blockchain having decentralization of information and other features like immutability can be used as a step to mitigate the implications. In this paper, a distributed e-tendering system based on IoT, smart contracts and hybrid cryptography is explored. The design consists of sections based on the way the tendering and bidding organization participates like creation and publication of tender by the tendering organization, bidding process by the bidder, evaluations of the proposed bid, selection of the best one to declare the winner and finally to track the development of the procurement process and give stakeholders real-time updates IoT devices are used. Maintaining transparency, traceability, fairness, and security is kept in mind throughout the process which is the main motive of this paper.
The Covid-19 (coronavirus) pandemic creates a worldwide health crisis. According to the WHO, the effective protection system is wearing a face mask in public places. Many studies proved that carrying a face mask is al...
详细信息
The identification of enzyme proteins is crucial for understanding the underlying mechanisms of various biological processes. In recent years, the development of machine learning techniques has made it possible to aut...
The identification of enzyme proteins is crucial for understanding the underlying mechanisms of various biological processes. In recent years, the development of machine learning techniques has made it possible to automate this task with high accuracy. In this study, we propose a method that uses 3-gram features and machine learning models for recognizing human enzyme protein sequences. First, the 3-gram features are extracted from the amino acid sequences of enzymes and use these features as input to several machine learning models, including naive bayes, decision tree, random forest, and Extra tree (ET). We compare the performance of these models using metrics such as accuracy, precision, recall, and F1 score. The results obtained show that the ET model performs the best, achieving an accuracy of 96%, precision of 93%, recall of 93%, and F1 score of 93%. Overall, the proposed approach demonstrates the potential of using machine learning models and 3-gram features for accurate identification of human enzyme protein sequences.
Text detection and classification play vital roles in various applications such as document analysis, scene text recognition, and image-based information retrieval. In recent years, deep learning approaches combined w...
Text detection and classification play vital roles in various applications such as document analysis, scene text recognition, and image-based information retrieval. In recent years, deep learning approaches combined with Optical Character Recognition (OCR) engines have shown promising results in text-related tasks. This research study investigates text detection and classification using popular OCR libraries, including Tesseract, Keras, Easy OCR, and Paddle OCR A comprehensive evaluation is conducted on a diverse dataset encompassing different text types, languages, and image qualities. Each OCR model’s performance is measured in terms of accuracy, precision, recall, and F1 score. The paper provides sample inputs and corresponding outputs for each model, allowing users to mention the strengths and limitations of each OCR approach. This study aims to assist the practitioners and researchers in selecting the most suitable OCR model for their specific text-related applications. The research findings demonstrate the potential and versatility of deep learning-based OCR solutions for text detection and classification tasks, paving the way for further advancements in this domain.
Sentiment Analysis is an essential process in the field of NLP (Natural Language Processing) that includes identifying the sentiment or emotion behind a text. Natural Language Processing (NLP) has a rapidly growing su...
Sentiment Analysis is an essential process in the field of NLP (Natural Language Processing) that includes identifying the sentiment or emotion behind a text. Natural Language Processing (NLP) has a rapidly growing subfield called sentiment analysis that aims to determine the sentiment or emotion underlying a given text. Using the TF/IDF (Term Frequency-Inverse Document Frequency) and Logical Regression techniques, In this study, we did a sentiment analysis on user reviews of products on Amazon. This study aims to assess the tone of Amazon customer reviews and give significant information into how customers see the items. A customer review dataset was acquired from kaggle and preprocessed to remove noise and extraneous information. Utilizing the TF/IDF technique, features were extracted from the preprocessed reviews. Then, these characteristics were utilised to train a Logistic Regression classifier to predict the reviews' sentiment. Standard performance indicators such as accuracy, In order to evaluate the performance of the classifier of Logistic Regression, precision, recall, & F1 score have been used.
In this work we investigate the problem of popular matching in a 3-uniform 3-partite hypergraph, where the first set contains a set of agents and the second and the third set contain different class of items respectiv...
In this work we investigate the problem of popular matching in a 3-uniform 3-partite hypergraph, where the first set contains a set of agents and the second and the third set contain different class of items respectively. Agents have a preference list for hyperedges, whereas the members of the other sets do not have preferences. Each agent votes in favor of a preferred matching. A matching M is called popular if there does not exist a matching M′ such that the number of agents that prefer M′ is more than the number of agents that prefer M. This paper provides a characterization of popular matching in a 3-uniform 3-partite hypergraph and shows that the problem of deciding whether a given 3-uniform 3-partite hypergraph has a popular matching is NP-hard. We also prove the NP-hardness of popular matching in a k-uniform k-partite hypergraph (k > 3). Assuming that a given 3-uniform 3-partite hypergraph admits at least one popular matching, we compare the cardinality of the maximum size popular matching (denoted as M ∗ ) with a maximum matching (denoted as M max ) in the 3-uniform 3-partite hypergraph and show that $\left| {{M^{\ast}}} \right| \geq \frac{2}{3}\left| {{M_{\max }}} \right|$.
Sentiment analysis is an important task in the field of natural language processing. The purpose of this article is to combine manually organized sentiment dictionaries with deep learning techniques to improve the acc...
详细信息
ISBN:
(数字)9798350359145
ISBN:
(纸本)9798350359152
Sentiment analysis is an important task in the field of natural language processing. The purpose of this article is to combine manually organized sentiment dictionaries with deep learning techniques to improve the accuracy of sentiment analysis. This article proposes a method based on improved feature vector for text sentiment analysis, which combines the semantic vector outputted by the deep learning model with the sentiment vector outputted by the emotional dictionary algorithm. Additionally, this article provides two combination schemes. The first solution is to combine in the sentence dimension. It involves calculating the sentiment score of the text through the emotional dictionary algorithm, converting it into a One-Hot vector, multiplying it with an Embedding matrix, and then converting it into a sentiment vector. The output of the deep learning model was concatenated with the semantic vector and then input into the decision layer. The second solution is to combine in the word dimension, filter out emotional words in the text, encode and preprocess emotional words, and assign a marker vector based on the type of emotional words. After combining the two vectors, input them into a deep learning model to generate an sentiment vector, and then concatenate them with the output vector of deep learning for decision-making. After validation through dataset experiments, the two models proposed in this article have demonstrated varying degrees of improvement across four key performance metrics: accuracy, precision, recall, and F1-score.
暂无评论