Search

Few studies have quantitatively analyzed the relationship among the step distance, the driving voltage, and driving frequency of the inchworm linear microactuator, which cannot achieve accurate control of the micro-actuator. What’s more, the parameter setting rule is not clear. In this paper, the relationship model among step distance, voltage, and frequency is studied through theoretical derivation and experimental analysis. The parameter setting conditions of the drive signal are analyzed, and the waveform is modified to improve the stability of the micro-actuator. The theoretical model of micro-driver step prediction is established by using the non-linear Support Vector Machine (SVM) regression theory and Gaussian radial basis kernel function and verified by experiments.

WhatsApp is a popular cross-platform application for communicating and connecting people. In this paper, the author analyzes massive data from WhatsApp public groups and obtains so-called tacit knowledge and social complex knowledge. During the processing, the author uses LDA(Latent Dirichlet Allocation) model to analyze the content which helps us to be aware of many aspects from public groups such as social trend, sentiment analysis and other hidden information. In the end, author get the results about topic for each group and the occurrence of words for each topic.

Instant message applications has become the trend of present social network for connecting people nowadays. With such massive information transmitted between end users, it may provide valuable hidden knowledge for data analysing and help us to obtain some insights of how the messages can influence current social trends. During the analyses, the author uses several different classifiers to conduct sentiment collection and use Ensemble Learning to aggregate the results. As a result, the author maps data about users’ sentiment graph in database of given category.