Aiming at the problems of fuzzy detection characteristics, high false positive rate and low accuracy traditional network intrusion technology, an improved intelligent method industrial Internet Things based on deep learning is proposed. Firstly, data set preprocessed transformed into 122 dimensional after one-hot coding; Secondly, aiming problem that convolution cannot deal with long-distance attributes, Bidirectional long short-term memory (BiLSTM) used to mine relationship between...

A novel asymmetric image encryption and authentication scheme based on equal modulus decomposition (EMD) in the Fresnel transform domain is proposed. First, spectrum of plaintext sparsely sampled. Then a sparse presentation divided into two complex-valued masks with EMD, both which are necessary for decryption authentication. The decrypted renders little information original visually. Furthermore, nonlinear correlation used to authenticate plaintext. In proposed here, illuminating wavelength...

We propose an asymmetric optical image encryption scheme with silhouette removal by using interference and equal modulus decomposition (EMD). Plaintext is first separated into two complex value masks the same EMD in Fresnel transform domain. The are encoded four phase-only (POMs), of which treated as ciphertexts other plaintext-dependent private keys inverse different diffraction distances interference-based encryption. Any information about plaintext, including its silhouette, cannot be...

Aiming at the problems of over reliance on labor and low generalization traditional emotion analysis methods based dictionary machine learning, an model microblog comment text deep learning is proposed. Firstly, obtained through crawler program. After data preprocessing, including cleaning, Chinese word segmentation, removal stop words, so on, Skip-gram used for vector training a large-scale unmarked corpus, then trained as input CNN-BiLSTM model, which combines Bidirectional Long-Short Term...

This article proposes a Weibo sentiment analysis method to improve traditional algorithms' efficiency and accuracy. The proposed algorithm uses deep learning in the Spark big data environment. First, input are converted into dynamic word vector representations using Chinese version of XLNet model. Then, dual-channel feature extraction is performed on TextCNN BiLSTM. an attention mechanism allocate computing resources efficiently realizes fusion classification. Comparative experiments...

To address the problems of high energy consumption and time delay offloading strategies in traditional edge computing, a computation strategy for Internet Things (IoT) using improved Particle Swarm Optimization (PSO) computing is proposed. First, system model an optimization objective function are constructed based on communication uplink transmission multiuser personalized task load while considering constraints from multiple aspects. Then, PSO used to update position particles by encoding...

In this article, the nonfragile passivity and passification problems are investigated for a class of nonlinear singular networked control systems (NCSs) with network‐induced time‐varying delay. particular, randomly occurring controller gain fluctuation is taken into consideration by introducing stochastic variable satisfying Bernoulli random distribution. By constructing proper Lyapunov–Krasovskii function, delay‐dependent sufficient conditions established to guarantee NCSs. Based on derived...

The method of computational intelligence to monitor and evaluate the concentration students in teaching process can promptly effectively adjust learning plan improve effect. In this article, clustering algorithm fuzzy control methods are used construct a research model students’ attention class. addition, article uses existing MATLAB-based image feature recognition detect obtain facial features analyze main expressions through techniques realize judgment attention. optimizes traditional...

To solve the problem of low positioning accuracy and ease environmental impact wearable devices in Internet things, a device indoor algorithm based on deep learning was proposed. Firstly, basic model composed an input layer, hidden output layer is proposed to realize continuous prediction with higher accuracy. Secondly, automatic stacking encoder trained signal strength data, features are extracted from large number samples noise build location fingerprint database. Finally, coding machine...

Abstract Aiming at the problem that users’ check-in interest preferences in social networks have complex time dependences, which leads to inaccurate point-of-interest (POI) recommendations, a location-based POI recommendation model using deep learning for network big data is proposed. First, original are fed into an embedding layer of dense vector representation and obtain user’s sequence (UCS) space-time interval information. Then, UCS spatiotemporal information sent bidirectional long-term...

Abstract With the continuous development and progress of society, information technology computer have also been highly developed improved. support technology, new technologies relay protection system emerged, which has great significance role for effective operation power system. Especially rural system, emergence brought many benefits, solved some problems in guaranteed effectiveness security This paper mainly analyses what is protection, specific application agricultural future trend protection.

The Computer Network course is widely offered in colleges and universities. This both theoretical practical. It organically combines theory practice order to achieve the goal of teaching. main objective train students have various skills, including those enterprise scale network construction, management, maintenance. Influenced by traditional teaching for a long time, existing universities urgently needs reform; otherwise, it will not be able meet social demand talents new era. paper focuses...

Facing the development of “Internet +” and demand diversified talents in market, excellent with computer operation ability are increasingly needed by all trades professions. This “ability” added to professional skills will enhance employment competitiveness a great extent. Therefore, traditional mode talent training need be transformed various integrated means adapt change status. By analyzing current situation agricultural application-oriented universities, fusion course is put forward this...