JOURNAL OF ARTIFICIAL INTELLIGENCE RESEARCH & ADVANCES

Complementary Approach of SVM for Heart Disease Prediction

Rajesh Yadav — 2023-11-03

To the point, Coronary-Heart-Disease is in records of killing huge number of lives every year. India contributes to 32% of deaths among all countries. In terms of data, 85% of them resulted from either heart attacks or strokes. In other terms, ⅘ people are risked to heart strokes. Over 75% of CHD deaths occur in low- and middle-income countries, where high blood pressure is one of the most important risk factors for middle age group to the risk of death through CHDs fall in 48 to 70. So, it is a need of time to focus on heart disease and make people in an lively active mode. To predict the accuracy in detection of heart disease we have attempted to describe the popular SVM machine learning along with its own set of strengths which is showed through complementary approach, Confusion Matrix. Our study includes following parts: Introduction, Factor affection heart functioning, Literature Review, SVM as Complementary approach, Methodology, Conclusion and References.

Diagnosing Pneumonia from Chest X-rays Using Deep Learning Algorithms Through Convolutional Neural Network, Transfer Learning and Fine Tuning

Kinjal Goswami — 2023-11-03

Pneumonia is an inflammatory condition of the lungs that induces air sacs which leads to a contagious infection of lungs. Patients who are afflicted with the virus can be saved from death and the virus can be eradicated from spreading further through effective diagnosis. X-rays of the chest are frequently used to diagnose pneumonia. Detecting pneumonia from a Chest X-ray is typically slow and inaccurate. It is essential to identify pneumonia quickly so that patients can receive prompt care, especially in rural areas. This work proposes a system that evaluates chest X-rays and categorizes the images using Deep Convolutional Neural Network Architecture, Transfer Learning and Fine Tuning on different CNN architectures. As part of this project, the implemented algorithms include CNN, VGG16, and Xception. Prior to building a model, Data Augmentation and Data Balancing is performed in order to improve the model’s generalization performance and accuracy. Among the above-mentioned algorithms, it has been deduced that VGG16 model has returned best accuracy.

Develop Institutional Chatbot Using Deep Neural Networks and NLTK

N. Sowjanya Kumari — 2023-11-03

Chatbots are intelligent software that can communicate and perform actions like those of a customer service representative. Chatbots are widely used for customer interaction and marketing on social networking and e-commerce sites. AI-based chatbots have the core ability to learn from any question based on initial training on a predefined dataset. A web-based platform provides a broad intelligent foundation for simulating human problem solving. The technology used here is based on deep neural networks and uses NLP for text processing and FLASK functionality for internal connectivity. Evolution has improved accuracy and performance rates on higher slopes. This recommended chatbot identifies the user context that triggers the specific intent of the response. Based on dynamic responses, it instantly generates the desired response for the user. The proposed system uses deep learning algorithms to train chatbots by experiencing different user responses and requests.

Fusion of Intelligence: Biosensors and AI in the Medical Landscape

Ushaa Eswaran — 2023-11-30

Healthcare stands poised to benefit immensely from emerging technologies like artificial intelligence (AI) and advanced biological sensors (biosensors). This study provides a comprehensive analysis on the integration of AI algorithms and biosensor devices for transformative health applications, assessing clinical relevance alongside ethical considerations. An overview of biosensors and AI in medical contexts is followed by examining how these technologies synergize for improved diagnostics, treatment personalization, remote monitoring, and other setups. Challenges around regulation, privacy, and AI bias are also highlighted. Case studies showcase cutting-edge implementations detecting cancer, monitoring neonatal health, and more, demonstrating evidence of clinical viability and tangible patient impact. Future directions centre on scalability enablers like biocompatible sensors, nanotechnology, and smart medical devices to deliver AI’s near-boundless analytical potential. The study concludes that multilayered datasets generated by AI-enabled biosensors can revolutionize medicine, but purposeful innovation and governance are vital to guide appropriate adoption while respecting ethical boundaries.

A Comparative Study on Various Machine Learning Techniques for the Prediction of Cardiac Ailment

Chandrani Chakravorty — 2023-10-31

Over the last couple of decades, cardiovascular complexities have become the leading source of death in impoverished regions. With heart attack rates on the acceleration at a youthful age, it is necessary to put in place a process to recognize the symptoms of a heart attack early and thus limit it. It is impossible for a common man to often undergo expensive tests such as an ECG and thus there must be a system that is both efficient and stable to forecast the tendency of heart illness. Early recognition of heart complications can diminish morbidity. However, it is impractical to explicitly supervise patients often and an expert’s consultation is not accessible as it needs knowledge and expertise. In this study we generated and analysed the models for forecasting heart illness, supporting a patient’s heart attributes, and detecting approaching cardiovascular illness using the techniques like Gradient Boost, AdaBoost, CatBoost etc., on a dataset accessible publicly on the Kaggle site, with the results significantly assessed using a confusion matrix. In contrast to other machine learning algorithms, the CatBoost classifier approach has the accuracy of 90.16%, in step with the trial conclusions.

A Systematic Review for Heart Disease Prediction Using AIML Algorithms

Kummari Jayasri — 2023-11-03

Cardiovascular diseases (CVDs), commonly known as heart diseases, have consistently held the position of being the primary global cause of mortality for many years. They are also the most serious illness in India and the rest of the world. So, a method that is reliable, accurate, and easy to use is needed to find these diseases early and start the right treatment. Using a variety of medical datasets, machine learning and deep learning techniques have been used to automate the analysis of large and complex data sets. In recent years, many researchers have used a wide range of methods to help doctors and other medical professionals find heart-related illnesses. This study looks at a number of models that were made using these methods and techniques, and it figures out how well they work. Researchers have a strong inclination towards models rooted in supervised learning techniques such as Support Vector Machines (SVM), K-Nearest Neighbour (KNN), Naive Bayes, Decision Trees (DT), Random Forest (RF), ensemble models, and various deep learning algorithms.

Ensuring Excellence: Quality Assessment of English Text and Speech Data Using NLP and ML

V. Laxmiprasad — 2023-11-03

The effectiveness and precision of speech processing and natural language processing (NLP) applications are substantially influenced by the quality of text and speech input. The combination of NLP and Machine Learning (ML) techniques has shown tremendous promise in recent years for improving the quality of written and verbal data. The present research enables a thorough investigation of techniques and strategies to enhance data quality using NLP and ML. The results discussed in this research provide useful insights into cutting-edge methods for combining NLP and ML to assess the quality of English data. The study promotes the reliability and efficacy of English data for various kinds of NLP and ML applications, such as sentiment analysis, information retrieval, and text categorization, and provides the foundation for additional research in this area.

Revolutionizing Product Marketing: Harnessing the Power of AI/ML

Lakshmi Namratha Vempaty — 2023-11-03

The rapid convergence of artificial intelligence (AI) and machine learning (ML) technologies in marketing has ushered in a new era of product promotion and engagement. This study delves into the many ways in which AI/ML can revolutionize product marketing strategies, showing a comprehensive range of methods and tools available. The presented high-level architecture illustrates the interconnectedness of AI-based marketing components, ranging from personalized recommendations and chatbots to sentiment analysis and video engagement prediction. A detailed exploration of AI/ML's potential is accompanied by a practical guide to employing these techniques. From personalization and predictive analysis to influencer collaboration and sentiment analysis, this study examines how each approach contributes to crafting effective marketing campaigns. Moreover, the integration of AI/ML in marketing is accompanied by a plethora of machine learning algorithms that facilitate various tasks. The provided taxonomy outlines a diverse range of algorithms, their purposes, and estimated timeframes for implementation.

A Secure Approach for Cross-chain Transactions Using Machine Learning Model

Madhuri Surisetty — 2023-10-31

The ability to conduct transactions or transfer assets between different blockchain networks is referred to as cross-chain transactions. It enables users to transfer assets from one blockchain network to another. In the Cryptocurrency ecosystem, the risk of fraudulent activities has become a significant concern. Due to these fraudulent activities, the cross-chain transactions have encountered challenges in terms of security and integrity. The need for robust fraud detection mechanisms becomes crucial to secure the integrity of transactions and protect investors from fraudulent activities. Detecting those fraudulent activities and preventing them is a challenging task due to the decentralized and pseudonymous nature of these digital assets. Machine Learning algorithms emerged as a power tool for fraud detection across various domains. This study proposes a novel secure approach for fraud detection in cryptocurrency transactions by leveraging machine learning algorithms. Our proposed methodology is to identify malicious activities and discern fraudulent transactions from legitimate ones. The machine learning models are trained on labelled datasets comprising both fraudulent and legitimate transactions by allowing them to learn patterns and detect anomalies indicative of fraudulent behavior. This proposed methodology contributes to the growing body of knowledge in cryptocurrency fraud detection and acts as a foundation for developing robust security measures and risk mitigation strategies to make the cross-chain transactions much stronger.

Stock Price Analysis and Forecasting Using Linear Regression and SVM Classifiers

Vinay Lowanshi — 2023-11-23

In this study, we attempt to implement a Machine Learning approach to predict stock market prices. Linear Regression is very effectively implemented in forecasting stock prices, returns, and stock modelling. This project is for common users as the prediction is done on all of the companies. We outline the design of the Linear Regression model with its salient features and customizable parameters. We select a certain group of parameters with a relatively significant impact on the share price of a company. With the help of analysis, the relation between the selected factors and share price is formulated which can help in forecasting accurate results. Although, share market can never be predicted, due to its vague domain, this project aims at applying Machine Learning in forecasting the stock prices.