Denoising of Images Using Deep Convolutional Autoencoders for Brain Tumor Classification

In the acquisition of images of the human body, medical imaging devices are crucial. The Magnetic Resonance Imaging (MRI) system detects tissue anomalies and tumours in the body of people. During the forming process, the MRI images are degraded by different kind of noises. It is difficult to remove certain noises, accompanied by the segmentation of images in order to classify anomalies. The most commonly explored areas of this period are automatic tumour detection systems using Magnetic Resonance Imaging. In the medical sector, timely and exact identification of frequencies is a problem. Automated systems are efficient that reduce human errors when tumour is detected. In recent years, many approaches have been proposed to do this, but there are still several drawbacks and a wide range of improvements on these methodologies are still needed. The image processing mechanism is widely used to improve early detection and treatment stages in the field of medical sciences. Sometimes the doctor can misdiagnose the image of MRI because of noise levels. To date, Deep Convolution Neural Networks (DCNN) have demonstrated excellent classification and segmentation efficiency. This paper proposes a technique for the image denoising using DCNN based Auto Encoders (DCNNAE) for achieving better accuracy rates in brain tumour prediction. In this paper we propose a deep convolution denoising auto encoder to remove noise from images and over fit the model problem by developing a deep convolution neural network for brain MRI image tumour prediction. The proposed model is compared with the existing methods and the results exhibits that the proposed model performance levels are better than the existing ones.

Firefly Optimization Based Noise Additive Privacy-Preserving Data Classification Technique to Predict Chronic Kidney Disease

With the continuous advancements in Information and Communication Technology, healthcare data is stored in the electronic forms and accessed remotely according to the requirements. However, there is a negative impact like unauthorized access, misuse, stealing of the data, which violates the privacy concern of patients. Sensitive information, if not protected, can become the basis for linkage attacks. Paper proposes an improved Privacy-Preserving Data Classification System for Chronic Kidney Disease dataset. Focus of the work is to predict the disease of patients’ while preventing the privacy breach of their sensitive information. To accomplish this goal, a metaheuristic Firefly Optimization Algorithm (FOA) is deployed for random noise generation (instead of fixed noise) and this noise is added to the least significant bits of sensitive data. Then, random forest classifier is applied on both original and perturbed dataset to predict the disease. Even after perturbation, technique preserves required significance of prediction results by maintaining the balance between utility and security of data. In order to validate the results, proposed method is compared with the existing technology on the basis of various evaluation parameters. Results show that proposed technique is suitable for healthcare applications where both privacy protection and accurate prediction are necessary conditions.

Corn Leaf Disease Detection with Pertinent Feature Selection Model Using Machine Learning Technique with Efficient Spot Tagging Model

Crop diseases constitute a substantial threat to food safety but, due to the lack of a critical basis, their rapid identification in many parts of the world is challenging. The development of accurate techniques in the field of image categorization based on leaves produced excellent results. Plant phenotyping for plant growth monitoring is an important aspect of plant characterization. Early detection of leaf diseases is crucial for efficient crop output in agriculture. Pests and diseases cause crop harm or destruction of a section of the plant, leading to lower food productivity. In addition, in a number of less-developed countries, awareness of pesticide management and control, as well as diseases, is limited. Some of the main reasons for decreasing food production are toxic diseases, poor disease control and extreme climate changes. The quality of farm crops may be influenced by bacterial spot, late blight, septoria and curved yellow leaf diseases. Because of automatic leaf disease classification systems, action is easy after leaf disease signs are detected. Applying image processing and machine learning methodologies, this research offers an efficient Spot Tagging Leaf Disease Detection with Pertinent Feature Selection Model using Machine Learning Technique (SPLDPFS-MLT). Different diseases deplete chlorophyll in leaves generating dark patches on the surface of the leaf. Machine learning algorithms can be used to identify image pre-processing, image segmentation, feature extraction and classification. Compared with traditional models, the proposed model shows that the model performance is better than those existing.

Optimization of the Convolution Operation to Accelerate Deep Neural Networks in FPGA

In recent years, machine learning algorithms related to images have been widely utilized by Convolution Neural Networks (CNN), and it has a high accuracy for recognition of an image. As CNN contains large number of computations, hardware accelerator like Field Programmable Gate Array is employed. Quite 90 % of operations during a CNN involves convolution. The objective of this work is to scale back the computation time to increase the peak, width and the pixel intensity levels in the input image. The execution time of a image processing program is mostly spent on loops. Loop optimization is a process of accelerating speed and reducing the overheads related to loops. It plays a crucial role in improving performance and making effective use of multiprocessing capabilities. Loop unrolling is one of the loop optimization techniques. In our work CNN with four levels of loop unrolling is used. Due to this delay is reduced compared with conventional Xilinix. With the assistance of strides and padding the 40 % of computation time has been reduced and is verified in MATLAB.

A Hybrid Model Integrating Singular Spectrum Analysis and Backpropagation Neural Network for Stock Price Forecasting

The primary purpose of trading in stock markets is to profit from buying and selling listed stocks. However, numerous factors can influence the stock prices, such as the company's present financial situation, news, rumor, macroeconomics, psychological, economic, political, and geopolitical factors. Consequently, tremendous challenges already exist in predicting noisy stock prices. This paper proposes a hybrid model integrating the singular spectrum analysis (SSA) and the backpropagation neural network (BPNN) to forecast daily closing prices in stock markets. The model first decomposes the stock prices into several components using the SSA. Then, the extracted components are utilized for training BPNNs to forecast future prices. Compared with the BPNN, the hybrid SSA-BPNN model demonstrates a better predictive performance, indicating the SSA's ability to extract hidden information and reduce the noise effect of the original time series.

A Transfer Learning Approach for Diabetic Retinopathy and Diabetic Macular Edema Severity Grading

Diabetic Retinopathy (DR) and Diabetic Macular Edema (DME) are complication that occurs in diabetic patient especially among working age group that leads to vision impairment problem and sometimes even permanent blindness. Early detection is very much needed for diagnosis and to reduce blindness or deterioration. The diagnosis phase of DR consumes more time, effort and cost when manually performed by ophthalmologists and more chances of misdiagnosis still there. Research community is working on to design computer aided diagnosis system for prior detection and for DR grading based on its severity. Ongoing researches in Artificial Intelligence (AI) have set out the advancement of deep learning technique which comes as a best technique to perform analysis and classification of medical images. In this paper, research is applied on Resnet50 model for classification of DR and DME based on its severity grading on public benchmark dataset. Transfer learning approach accomplishes the best outcome on Indian Diabetic Retinopathy Image Dataset (IDRiD).

Hate Speech in the Arab Electronic Press and Social Networks

Nowadays we are witnessing an open world, characterized by globalization which is accompanied by a technology through which information circulates without borders, especially with the widespread use of social networking sites being the most common communication tool, that gives access through various applications to a large space for the presentation of multiple ideas, including extremist ideas, and the spread of hate speech. This paper introduces a system of detection of hate speech in the texts of Arabic read media and social media, which is based on a combined use of NLP, and machine learning methods. The training of the detection model is done on a large Dataset of articles, tweets and comments, collected, balanced and tokenized afterwards using BERT in Arabic. The trained model detects hate speech in Arabic and various Arabic based dialects, by classifying the texts into two classes: Neutral and Abusive. The above-mentioned model is evaluated using precision metrics, recall and f1 score, it has reached an accuracy of 83%.

Vigorous IDS on Nefarious Operations and Threat Analysis Using Ensemble Machine Learning

The geometric increase in the usage of computer networking activities poses problems with the management of network normal operations. These issues had drawn the attention of network security researchers to introduce different kinds of intrusion detection systems (IDS) which monitor data flow in a network for unwanted and illicit operations. The violation of security policies with nefarious motive is what is known as intrusion. The IDS therefore examine traffic passing through networked systems checking for nefarious operations and threats, which then sends warnings if any of these malicious activities are detected. There are 2 types of detection of malicious activities, misuse detection, in this case the information about the passing network traffic is gathered, analyzed, which is then compared with the stored predefined signatures. The other type of detection is the Anomaly detection which is detecting all network activities that deviates from regular user operations. Several researchers have done various works on IDS in which they employed different machine learning (ML), evaluating their work on various datasets. In this paper, an efficient IDS is built using Ensemble machine learning algorithms which is evaluated on CIC-IDS2017, an updated dataset that contains most recent attacks. The results obtained show a great increase in the rate of detection, increase in accuracy as well as reduction in the false positive rates (FPR).

Automatic Short Answer Grading System in Indonesian Language Using BERT Machine Learning

A system capable of automatically grading short answers is a very useful tool. The system can be created using machine learning algorithms. In this study, a machine system using BERT is proposed. BERT is an open-source system that is set to English by default. The use of languages other than English Language is a challenge to be implemented in BERT. This study proposes a novel system to implement Indonesian Language in the BERT system for automatic grading of short answers. The experimental results were measured using two measuring instruments: Cohen's Kappa coefficient and the Confusion Matrix. The result of measuring the BERT output of the implemented system has a Cohen Kappa coefficient of 0.75, a precision of 0.94, a recall of 0.96, a Specificity of 0.76 and an F1 Score of 0.95. Based on the measurement results, it can be seen that the implementation of the automatic short answer grading system in Indonesian Language using BERT machine learning has been successful.

Efficient Computation for Localization and Navigation System for a Differential Drive Mobile Robot in Indoor and Outdoor Environments

Numerous challenges are usually faced during the design and development of an autonomous mobile robot. Path planning and navigation are two significant areas in the control of autonomous mobile robots. The computation of odometry plays a major role in developing navigation systems. This research aims to develop an effective method for the computation of odometry using low-cost sensors, in the differential drive mobile robot. The controller acquires the localization of the robot and guides the path to reach the required target position using the calculated odometry and its created new two-dimensional mapping. The proposed method enables the determination of the global position of the robot through odometry calibration within the indoor and outdoor environment using Graphical Simulation software.