الفهرس | Only 14 pages are availabe for public view |
Abstract In this thesis, we proposed a model that can work as a truthfulness classifier based for IoT Big Data. The proposed model operates in two modes: preprocessing and processing. The preprocessingmode converges on reducing the factors that hinder making efficient processing by focusing on three main stages. First, handling the existence of missing data is by applying interval-valued fuzzy-rough (IVFR) feature selection methodology. It highlights the most important features that contain missing data and gets rid of the others. Then, Maximum Likelihood (ML) approach is used for estimating the missing values. Second, anomalies are detected by initially utilizing K-Nearest Neighbors (KNN) algorithm then removing the detected ones from the data. Third, the dimensionality of nonlinearly separable data is reduced by exploiting Self-Organizing Map (SOM) network. The processing mode enables passing the prepared data to a straightforward classifier based on a Deep Learning approach. We make use of autoencoder networks in constructing a deep network. The proposed Deep Stacked AutoEncoder (DSAE) consists of a stack of autoencoders accompanying with a softmax layer. The extracted features by the DSAE are non-handcrafted and task dependent, which gives it the most discriminative power to work as an efficient Deep Learning classifier for representing a high-level abstraction of IoT Big Data. We applied the proposed model to the PAMP2 Physical Activity Monitoring dataset. It is a large scale sensor dataset and can be taken as an example for IoT Big Data. Experimental results show that DSAE achieves high accuracy (99.8%) compared to the state-of-the-art shallow classifiers. |