Search

In healthcare field, Alzheimer's disease (AD) a neurodegenerative issue wherein the demise of synapses causes disarray, strain to recall and intellectual decay. Alzheimer disease mostly affects older adults and in rare case it affects the young ones. Brain images used primarily and mostly to detect the Alzheimer's disease by Magnetic resonance imaging (MRI). A contrast material used here to improve analysis of disease with help of chemical substance called Gadolinium contrast media. Dementia syndrome, a most common cause of Alzheimer's due to damage of brain cells or degeneration of brain cells. In proposed methodology, features of brain tissue shrinkage has been taken for primitive diagnosis of Alzheimer's disease. Gadolinium material used to improve identification of affected area in the brain region. Various machine learning and Deep Learning techniques are implemented, and comparative analysis of performance metrics has carried out for detecting AD in this research. Deep learning algorithm considered as a better solution provider for this problem (i.e. it helps us to identify the disease early with good accuracy when compared to machine learning algorithm).

Smart IOT-based devices are used to extract the emotions and sentiment metrics expressed by the course members for calculating the utility function value for each e-learning content. This is mainly used for rating the E-learning content available on online sources. ELMs are used to combine the output of classifier models implemented by CNN and RNN. Text, image, and acoustic data with respect to time are extracted by CNN, and the extracted output is processed by both CNN and RNN for sentiment analysis. The hidden layers of CNN are trained with ensemble-based algorithms, namely EN-ELM and progressive EN-ELM. The output of CNN and RNN is fed to the two consecutive ELM further whose output is fed to the SVM classifier. Unlike other CNN algorithms, ensemble algorithms are used in the hidden neurons of CNN to carry the time series information to the RNN with the help of stumps and identity activation function. The accuracy of progressive convolution neural network is compared to the accuracy of CNN with RNN. Arousal in the accuracy level is detected with 93–97%.

Multimodal sentiment analysis framework is tremendously applied in the medical and healthcare sectors. Identifying the depression, differently disabled person speech recognition, Alzheimer, low pressure or heart problems and those kinds of impairments are widely addressed. Video data is processed for mining feature polarity from acoustic, linguistic and visual data upon extraction from the same. The feature data set extracted from the YouTube videos contains comments, likes, views and shares for expressing the polarity of information conveyed through the streaming videos. Static information from a video file is extracted in the form of linguistic representation. Musical data extracted and transformed in the linguistic form is used for polarity classification using ensemble-based random forest algorithm which has encountered with the error rate of 4.76%. Short feature vectors are expressed in the visualizing musical data and trending YouTube videos data set for utilizing the transformed and SF vectors from video and musical data. Accuracy of the ensemble-based learning is obtained as 91.6% which is tougher than any other algorithms to achieve using the same set of machine learning algorithms. Proper filter wrapping of batch data is used for the split ratio of 5 percentage split window. When SVM is used in discrimination with ensemble random forest algorithm, the predicted results contain an error rate of 2.64% which improves the accuracy of the classifier along the soft margin with an accuracy of 96.53% accuracy.

Word Sense Disambiguation refers to the task of correctly identifying the sense of an ambiguous word in a given context. This is considered as an AI-complete problem in the field of natural language processing. In this paper, an LDA-based WSD model is proposed which, to the best of our knowledge, is the first work that makes use of LDA for disambiguation in Malayalam. We had selected three ambiguous words and manually collected sentences including those from different novels, short stories, and web documents. The corpus consists of more than 200 instances of each ambiguous word. The model performed well compared to the state-of-the-art WSD systems in Malayalam.

Customer churn is an important issue and major concern for many companies. This trend is more noticeable in Telecom field. Telecom operators requires an essential proactive method to prevent customer churn. The existing works fails to adopt best feature selection for designing model. This works contributes on developing churn prediction model, which helps telecom operators to identify the customers who are about to churn. The significance for the recall evaluation measure, which actually solves the real-time business problem is highlighted. The prominent goal of this churn analysis is to perform binary classification with customer records and figure out who are likely to cancelled in the future. Fifteen machine learning methods with different parameters are employed. The performance of the model is evaluated by various measures like Accuracy, Recall, Precision, F-score, ROC and AUC. Our aim in this work is to produce highest recall value which has a direct impact on real-world business problems. Based on experimental analysis, we observed that Decision Tree model 3 outperforms all other models.

Electoral fraud or election fraud is an illegal intervention in the process of election, in which voters vote repeatedly for a favored party thereby increasing vote share. In Order to conduct elections ethically, this election fraud should be eliminated. Hence, the automatic voting system using Convolutional Neural Network (CNN) has been proposed. The Existing systems are time-consuming and difficult-to-maintain since they include manual processes. Presently different kinds of biometric processes available, among that face detection is the most effective. The proposed system develops an automatic voting system without any human intervence. This system includes a camera for capturing the images of electors, captured images will be stored in the database for analysis. data analysis will be performed on the database. In this method, all the label images are trained through the convolutional neural network to predict the output by classifying the images. This classification system produces an accuracy of approximately 90%.

Using facial recognition technology, provide the best security system with alerts for smart devices such as laptops and PCs. The design descriptions of the various modules involved in automatic face recognition are presented in this paper: face detection, face identification, and face verification. Also we’re visiting further process the image captured by CV to test whether the person is allowed or unauthorized to access the laptop or PC using openCV and native Binary Pattern Histogram (BPH) Algorithm. The LBPH algorithm is also a simple solution to the problem of face recognition, since it can identify both the front and side faces. A updated LBPH algorithm assisted pixel neighbourhood grey median (MLBPH) is proposed to solve this problem. After detection, the image is verified and allows only the user to access the laptop or PC if he/she’s authorized. And if not, an e-mail or message is pushed to the respective person to alert that somebody else is using his/her laptop by enhance the privacy.

Due to this recent wave of the Deadly corona virus disease, which has spread widely all over the world, so physical or Social distancing has become a mandatory precaution to avoid close contact. As of January 2021, the count of affected cases is 10.8 million and recovered cases are 10.5 million and the death rate is 155 thousand. In conformity, we have proposed a system that elucidates the usage of Python, computer vision, and deep learning to monitor whether people are maintaining Social distance at public areas in roads, pathways, work places and, even in hospitals. For the purpose of monitoring social distances among people, we have developed a social distancing detection tool that can keep track of safe distance and determine the distance between people in order to reduce the impact of the corona virus pandemic by analyzing real-time video Streaming, such as CCTV or Camera which can be added with the security camera system at hospitals, workplaces, markets, Jewelers shop etc. As an output, the detection tool will throw A Voice alert or a Notification, so that people would be remained to maintain a social distance.

Images consist of visual components such as color, shape, and texture. These components stand as the primary basis with which images are distinguished. A content-based image retrieval system extracts these primary features of an image and checks the similarity of the extracted features with those of the image given by the user. A group of images similar to the query image fed is obtained as a result. This paper proposes a new methodology for image retrieval using the local descriptors of an image in combination with one another. HSV histogram, Color moments, Color auto correlogram, Histogram of Oriented Gradients, and Wavelet transform are used to form the feature descriptor. In this work, it is found that a combination of all these features produces promising results that supersede previous research. Supervised learning algorithm, SVM is used for classification of the images. Wang dataset is used to evaluate the proposed system.

The paper proposes FRECCA algorithm to find out the clusters of sentences having the inter relation among them. The system uses the fuzzy clustering algorithm that compares the sentences and find out the similarity value and forms a cluster. The fuzzy clustering algorithm checks the possibility of the sentence to which cluster it belongs to. In this paper a Genetic algorithm approach is also suggested that checks the highest similarity value between the sentences and grouped to form a cluster in a more sophisticated manner, there by forming n number of clusters from n sentences.