Search

With the increased use of social network sites, such as Twitter, attackers exploit these platforms to spread counterfeit content. Such content can be fake advertisements or illegal content. Classifying such content is a challenging task, especially in Arabic. The Arabic language has a complex structure and makes classification tasks more difficult. This paper presents an approach to classifying Arabic tweets using classical machine learning (non-deep machine learning) and deep learning techniques. Tweets corpus were collected through Twitter API and labelled manually to get a reliable dataset. For an efficient classifier, feature extraction is applied to the corpus dataset. Then, two learning techniques are used for each feature extraction technique on the created dataset using N-gram models (uni-gram, bi-gram, and char-gram). The applied classical machine learning algorithms are support vector machines, neural networks, logistics regression, and naïve Bayes. Global vector (GloVe) and fastText learning models are utilised for the deep learning approaches. The Precision, Recall, and F1-score are the suggested performance measures calculated in this paper. Afterwards, the dataset is increased using the synthetic minority oversampling technique class to create a balanced dataset. After applying the classical machine learning models, the experimental results show that the neural network algorithm outperforms the other algorithms. Moreover, the GloVe outperforms the fastText model for the deep learning approach. [ABSTRACT FROM AUTHOR]

Learning from a class-imbalanced data poses a challenging task in machine learning. Because the skewness in imbalanced data is inherent; therefore, the number of instances is not equal. The class with few samples is identified as the minority class, and a class with many instances is referred to as the majority class. Due to this, biased learning is caused by the machine learning models. When the dataset had more features, the classifier’s performance also suffered. A common method for choosing the crucial features from a dataset and enhancing the performance of a machine learning classifier is feature selection. The feature selection cannot be handled by oversampling approaches alone. Therefore, the purpose of the suggested strategy is to enhance deep learning classifier performance by feature selection-based oversampling. The three modules that make up the suggested methodology are deep 1D-convolutional neural network (DCNN) for classification, SMOTE for oversampling, and genetic algorithm for feature selection. High-dimensional datasets with class imbalance issues are considered by the proposed method. There are both small and large datasets used. In addition, the datasets with low and large imbalance ratios are considered. When compared to benchmark methods, the suggested solution performs well. [ABSTRACT FROM AUTHOR]

As a widely known disease diabetes mellitus makes the human body produce quite less hormone and also tend to cause increased glucose that results in abnormal metabolism of varied organs in the body like eyes, kidneys, etc. Diabetic analysis has attracted the research community to treat some missing values and class imbalance issues. The performance of diabetes mellitus classification by the usage of machine learning techniques is comparatively low. We suggest this paper on imbalanced dataset with missing values, an efficient prediction algorithm for diabetes mellitus classification using Deep 1D-Convolutional Neural Network values. The outlier detection is used for removing missing values first. Then, oversampling method (SMOTE) is used to reduce the influence of imbalance class on prediction performance. Finally, predictions are produced using a DCNN classifier and are evaluated using a selective set of evaluation indicators. Experiments on the Pima Indian diabetes dataset (PIDD) from UCI Repository (University of California at Irvine) have yielded positive results. Our proposed DCNN algorithm has been shown to be successful and superior. [ABSTRACT FROM AUTHOR]

GPGPUs no longer seem to be an inconsequential component of supercomputing architectures, and a section of HPC application developers no longer refrain from utilizing GPGPUs. CUDA, in general, has remained a successful computing platform for those architectures. Thousands of scientific applications from various domains, such, as bio-informatics, HEP, and so forth, have been accelerated using CUDA in the past few years. In fact, the energy consumption issue still remains a serious challenge for the HPC and GPGPU communities. This paper proposes energy prediction approaches using dynamic regression models, such as parallel dynamic random forest modeling (P-DynRFM), dynamic random forest modeling (DynRFM), dynamic support vector machines (DynSVM), and dynamic linear regression modeling (DynLRM). These models identify energy efficient CUDA application instances while considering the block size, grid size, and the other tunable parameters, such as problem size. The predictions of CUDA application instances have been attained by executing a few CUDA application instances and predicting the other CUDA application instances based on the performance metrics of applications, such as number of instructions, memory issues, and so forth. The proposed energy prediction mechanisms were evaluated with CUDA applications such as Nbody and Particle Simulations on two GPGPU machines. The proposed dynamic prediction mechanisms achieved a 50.26 to 61.23 percentage of energy/performance prediction improvements when compared to the classical prediction models; and, the parallel implementation of the dynamic RFM (P-DynRFM) recorded over 83 percentage points of prediction time improvements. [ABSTRACT FROM AUTHOR]

The Himalayas has one of the largest concentrations of glaciers outside the Polar Regions. Various reports suggest that significant number of mountain glaciers is shrinking due to climatic variations. Monitoring of these glaciers is important to assess future availability of water resources in the Himalayan region. However, Himalayan glaciers are normally difficult to monitor due to the rugged, mountainous terrain. Therefore, images of Indian Remote Sensing Satellite were used to monitor glaciers in the Baspa basin. Investigations have shown the presence of 30 glaciers in the basin, with areal extent of 167 km2. Out of these, 19 glaciers, with areal extent of 140 km2 were selected to estimate retreat. Investigation suggests that almost all glaciers are retreating in the study basin and overall 19% deglaciation has been observed from 1962 to 2001. In general, altitude distribution appears to have significant influence on glacial retreat. Glaciers located around 5000 m altitude range are showing 24% loss as compared to 14% by glaciers located in altitude range higher than 5400 m. In addition, mean altitude of glacier terminus is shifted upward by 88 m, i.e. from 4482 to 4570 m in last 39 years. The glacial volumes were estimated using regression relationship between area and depth. The investigations have suggested that 19.10 km3 of glacial water stored in the 19 glaciers in 1962, has been reduced to 14.71 km3 in 2001, respectively, an overall loss of 23 percent in a period between 1962 and 2001. These investigations suggest that all glaciers in the Baspa Basin are reducing and in long term, such reducing trend can create scarcity of water in the region. [ABSTRACT FROM AUTHOR]