Medical Multimedia Big Data Analysis Modeling Based on DBN Algorithm

Abstract

With the development of medical multimedia analysis methods based on DBN, DBN models have gained the ability to surpass medical experts in the evaluation of multimedia in some clinical examinations. Firstly, based on the existing architecture of the Internet of Things, combined with the actual characteristics of the hospital, the medical multimedia data is accessed from the IoT support platform. Secondly, the medical multimedia data modeling and classification method based on DBN is studied and analyzed. Three network structure models, a deep belief network, a stacking automatic encoder and a convolutional neural network, were introduced and analyzed. The medical multimedia data classification modeling method based on DBN was proposed to further improve the accuracy of medical multimedia data classification. The experimental results show that compared with the traditional feature extraction based neural network classification method, the classification performance is better. Thirdly, the medical state assessment model is constructed based on the multivariate Gaussian distribution theory. To study how to use the multivariate Gaussian distribution theory to design an evaluation model that can evaluate the health status of users efficiently and accurately. Finally, using the MATLAB software platform, through the experiment and simulation of 40 groups of 8×8064-dimensional physiological big data of 32 volunteers, First, determine the optimal parameters of a set of health assessment models; then use the model to learn the characteristics of physiological parameters; finally, the state assessment model to obtain the health assessment results. The experimental results show that the feature learning model based on convolutional neural network theory can effectively extract the deep features of medical multimedia big data. The health state assessment model based on multivariate Gaussian distribution theory can effectively evaluate the health status of human body. INDEX TERMS DBN, medical multimedia, neural network, feature extraction, multivariate Gaussian distribution.

Sensor Network Security Risk Prediction and Control Method Based on Big Data Analysis

“…e sampling carrier frequency time width of FBG sensor network information transmission risk is 1.25 ms, the normalized initial frequency of risk prediction is f 11 � 0.05, f 12 � 0.15, the number of sampling points of risk information data is N � 120, and the variation range of interference signal-to-noise ratio is −12∼12 dB. According to the above simulation environment and parameter settings, other algorithms and this algorithm are selected to collect the risk data of wireless sensor networks, respectively, and take the collected data as the test object to simulate the dynamic prediction of information transmission risk of high reliability fiber Bragg grating sensor networks [24,25], and the comparison of time cost of information transmission risk prediction of high reliability fiber Bragg grating sensor network is shown in Figure 3.…”

A new method for predicting and controlling security risk in sensor networks using big data analysis

“…In light of the aforementioned, the danger information from wireless sensor networks is used in this method as well as other algorithms, which are selected for the simulation environment and parameter settings. high-performance fiber the dynamic prediction of information transmission risk was modelled using Bragg grating sensor networks [23,24] (Figure3). High-reliability fib grating sensor networks' information transmission time and risk assessment are also assessed.…”

Single Pole‐to‐Ground Fault Location Method for MMC‐HVDC System Using Wavelet Decomposition and DBN

“…DBN is a generative model. By training the weights between its neurons, the entire neural network will generate training data based on the maximum probability [22]. The model is based on the restricted Boltzmann machine (RBM) and a multihidden layer neural network which is composed of multiple RBM stacks.…”