Accurate and quick brain tumor diagnosis and treatment are critical for the successful management of these disorders, which can have a substantial influence on an individual’s quality of life and even be deadly. This work describes a novel technique for brain tumor segmentation and classification that employs three deep learning architectures: GlobalNet, Multi-task Learning, and FusionNet. GlobalNet is a convolutional neural network that incorporates global context information into the model via a global pooling layer, whereas Multi-task Learning allows the model to do numerous tasks concurrently, enhancing each task’s performance by using shared knowledge among them. FusionNet is a deep fully residual convolutional neural network that combines information extracted from multiple CNNs trained on various medical imaging modalities to improve the accuracy and resilience of brain tumor segmentation. The proposed method is shown to be robust and accurate in segmenting and categorizing brain tumors, providing an effective solution with better accuracy in predicting the brain tumors and segmenting the affected part

S. Shreyanth, S. Niveditha and V. Kathiroli, "Accurate Brain Tumor Segmentation and Detection using Multi-Task Learning with GlobalNet and FusionNet," 2023 IEEE 12th International Conference on Communication Systems and Network Technologies (CSNT), Bhopal, India, 2023, pp. 478-485, doi: 10.1109/CSNT57126.2023.10134722.

Speech signal segmentation is a key component of speech processing tasks, such as speaker recognition, speech recognition, and emotion recognition. Traditional approaches often have low accuracy and high computational complexity. To address this issue, we propose a novel approach using kernelized deep networks. The method utilizes Mel-Frequency Cepstral Coefficients (MFCC) and Linear Predictive Coding (LPC) for feature extraction, with Spectral subtraction, wavelet denoising and Wiener filtering as preprocessing techniques. Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA) are used for dimensionality reduction. Spectral clustering using the graph Laplacian and normalized Laplacian matrix is then employed to cluster the speech signals into different segments. The Lanczos algorithm, Implicitly Restarted Arnoldi Method and Krylov subspace is used to train the networks and classify the signals. Results on TIMIT and BURN corpus demonstrate that the proposed approach outperforms traditional techniques in terms of segmentation accuracy and reduces computational complexity. This research highlights the potential of kernelized deep networks for speech segmentation and contributes to the field of speech processing.

S. Niveditha, S. Shreyanth, V. Kathiroli, P. Agarwal and S. Ram Abishek, "Kernelized Deep Networks for Speech Signal Segmentation Using Clustering and Artificial Intelligence in Neural Networks," 2023 IEEE 12th International Conference on Communication Systems and Network Technologies (CSNT), Bhopal, India, 2023, pp. 667-674, doi: 10.1109/CSNT57126.2023.10134609.

Grid computing based on clusters has emerged as a promising strategy for improving the efficacy of wireless network data transmission. This study examines the incorporation of cluster-based grid computing, routing analysis protocols, and deep learning techniques to optimize data transmission in wireless networks. The proposed method utilizes clusters to distribute computing duties and enhance resource utilization, resulting in efficient data transmission. To further improve the routing process, a novel routing analysis protocol is introduced, which dynamically adapts to network conditions and chooses the most optimal routes. In addition, deep learning algorithms are used to analyze network data patterns, allowing for intelligent data routing and resource allocation decisions. Experiment results exhibit the efficacy of the proposed method, revealing substantial enhancements in network performance metrics such as throughput, latency, and energy consumption. This research contributes to the development of cluster-based grid computing and offers valuable insights for the design of efficient wireless network data transmission systems.

S., Shreyanth & S., Niveditha. (2023). CLUSTER-BASED GRID COMPUTING ON WIRELESS NETWORK DATA TRANSMISSION WITH ROUTING ANALYSIS PROTOCOL AND DEEP LEARNING. International Journal of Advanced Research. 11. 517-534. 10.21474/IJAR01/17096.

The internet is saturated with images that convey messages and emotions more effectively than words alone in today's digital age. Individuals with visual impairments, who are unable to perceive and comprehend these images, face significant obstacles in this visual-centric online environment. As there are millions of visually impaired people around the globe, it is essential to close this accessibility gap and enable them to interact with online visual content. We propose a novel model for neural image caption generation with visual attention to address this pressing issue. Our model uses a combination of CNNs and RNNs to convert the content of images into aural descriptions, making them accessible to the visually impaired. The primary objective of our project is to generate captions that accurately and effectively describe the visual elements of an image. The model proposed operates in two phases. First, a text-to-speech API is utilized to convert the image's content into a textual description. The extracted textual description is then converted to audio, allowing visually impaired individuals to perceive visual information through sound. Through exhaustive experimentation and evaluation, we intend to achieve a high level of precision and descriptivism in our system for image captioning. We will evaluate the performance of the model by undertaking comprehensive qualitative and quantitative assessments, comparing its generated captions to ground truth captions annotated by humans. By enabling visually impaired individuals to access and comprehend online images, our research promotes digital inclusion and equality. It has the potential to improve the online experience for millions of visually impaired people, enabling them to interact with visual content and enriching their lives through meaningful image-based interactions.

Priyanka Agarwal, Niveditha S, Shreyanth S, Sarveshwaran R, Rajesh P K, " Neural Image Caption Generation with Visual Attention : Enabling Image Accessibility for the Visually Impaired, International Journal of Scientific Research in Science, Engineering and Technology(IJSRSET), Print ISSN : 2395-1990, Online ISSN : 2394-4099, Volume 10, Issue 3, pp.562-575, May-June-2023.

Machine learning has made an impact on the area attributed to microchip, and it is initially used in automation. These techniques will eventually supplant the current VLSI design concept. Design creation has been automated by substituting time-consuming traditional concepts developed by experts. This development could result in a tremendous change in the realm of hardware computation and AI’s powerful analysis tools. As a result, during the last four decades, several tasks have been computerized, and a plethora of sophisticated tasks have been mechanized. And then if someone has invented a new concept (in terms of computing, analyzing, optimization, and inter-relationship manufacturing) and the invention process is computerized, major firms like IBM and Intel have their own CAD division to handle these issues for design automation. Several firms, including major electronic design automation company, sell CAD software’s based on the employment of intelligent tools at circuit development. ML had broadened the present reach by assisting with plausible solutions for an extensive kind on the subject problems as well as challenges in a different range of technical domains. Importance of machine learning in the EDA solutions industry has increased its capabilities by lowering the man-hours spent on design confirmation and its implementation, cost reductions, and design product productivity. In this article we have discussed the importance of the application of machine learning in VLSI chip design and development and how we implemented ML-oriented BIST.

Shreyanth, S., Harshitha, D.S. & Niveditha, S. Implementation of Machine Learning in VLSI Integrated Circuit Design. SN COMPUT. SCI. 4, 137 (2023).

Fintech, or financial technology, has revolutionized the way consumers access and manage their finances. In times of crisis, retaining customers becomes even more important for financial firms to maintain market share and minimize financial losses. This research aims to explore how technology-driven financial firms adapt their marketing strategies to retain customers during times of crisis. Through a literature review and case studies, this study will examine the methods used by fintech companies to maintain customer loyalty and engagement during economic downturns. It will also examine the tactics, techniques and models used by fintech companies to retain customers, with a focus on the role of financial risk management, CRM systems and customer data analysis. The paper will also discuss the importance of customer retention in the context of the fintech industry and its impact on financial stability. It will also delve into the evaluation methods and models used by the firms like the Pareto principle, Kano model, RFM analysis and Entropy-weighted CLV for identifying and prioritizing their most valuable customers and tailoring their retention strategies accordingly. Additionally, this research will investigate the role of digital marketing channels, such as social media and email marketing, in fintech crisis management. The findings of this study will provide valuable insights for fintech companies, as well as marketers and practitioners in the financial industry, on how to effectively navigate crisis situations and maintain customer relationships.

Shreyanth, S., Suwetha, P. T., Kathiroli, V., Niveditha, S., & Jayaprakash, H. (2023). Fintech, Crisis, and Marketing: How Technology-Driven Financial Firms Adapt Their Approach to Retain Customers. Proceedings of the International Conference on Emerging Trends in Business & Management (ICETBM 2023), 309–321. Atlantis Press. https://doi.org/10.2991/978-94-6463-162-3_28

Cerebral vascular problems, such as stroke, have become a major concern in healthcare due to their impact on neurological health. Early detection of these illnesses can considerably enhance patient outcomes and lead proper treatment choices. Recently, supervised machine learning algorithms have emerged as potential techniques for anticipating cerebrovascular diseases utilizing numerous clinical indicators. This study compares the predictive performance of three supervised learning algorithms, Random Forest, Decision Tree, and XGBoost, in the context of cerebrovascular disease prediction. The dataset utilized for training and evaluation comprises a variety of clinical parameters such as age, blood pressure, cholesterol levels, smoking status, and medical history. Random Forest has the highest prediction accuracy of the three algorithms, at 99.93% for cerebrovascular diseases. This result demonstrates how effective and reliable Random Forest’s pattern detection algorithms are in this dataset. With a near accuracy of 97.94%, XGBoost exhibits a good capacity for prediction. Decision Tree nevertheless obtains an excellent accuracy of 97.53% despite somewhat inferior performance compared to XGBoost, making it a competitive choice for the prediction of cerebrovascular disease. These results outperform the most recent state-of-the-art results and show a considerable increase in accuracy when compared to existing methods. The study emphasizes how well supervised learning systems, in particular Random Forest, capture the complex relationships between clinical parameters for accurate predictions of cerebrovascular diseases. The outstanding accuracy of these models highlights their potential for accurate cerebrovascular prediction in clinical practice.

B. Sethuraman and S. Niveditha, "Cerebrovascular Accident Prognosis using Supervised Machine Learning Algorithms," 2023 World Conference on Communication & Computing (WCONF), RAIPUR, India, 2023, pp. 1-8, doi: 10.1109/WCONF58270.2023.10235122.

Malware's increasing menace in the digital realm needs the development of powerful detection and classification systems. This study presents a unique method for predicting malware category and family using machine learning, leveraging the Cuckoo environment and automated feature selection. To undertake an exhaustive examination of malware activity, we combined the Cuckoo environment with cutting-edge machine learning methods, such as the k-Nearest Neighbors (KNN) technique. The dynamic analysis of malware samples within the Cuckoo environment, which captured their interactions with the execution environment, provided remarkable insights into their malicious behaviors. The Boruta method aided automated feature selection, improving the feature set and optimizing model performance. A comparison with existing models yielded striking findings. Notably, Test Case 3 surpassed earlier cases by incorporating Automated Feature Selection + Cross-Validation. In identifying phishing attempts, it displayed outstanding specificity (90%), precision (93%), recall (96%), and an impressive F1-Score (92%), indicating its proficiency in reliably recognizing this frequent threat. Test Case 3 showed considerable increases as well, with a surprising 7.0% increase in precision and a notable 14.6% increase in recall when compared to Test Case 1. Furthermore, in the field of ransomware detection, Test Case 4, which solely focused on Automated Feature Selection, obtained excellent results, with 91% specificity, 92% accuracy, and a recall rate of 93%. These developments highlight the importance of the Boruta algorithm in optimizing the model's performance.

Niveditha S & Rr, Prianka & K, Sathya & S, Shreyanth & Subramani, Nandhagopal & Deivasigamani, Balakrishnan & S, Karthikeyan. (2024). Predicting Malware Classification and Family using Machine Learning: A Cuckoo Environment Approach with Automated Feature Selection. Procedia Computer Science. 235. 2434-2451. 10.1016/j.procs.2024.04.230.

Finding the appropriate candidate for a position can be a difficult and time-consuming effort. The sheer volume of resumes and an expanding applicant pool makes it difficult for hiring managers to find the best-suited applicants quickly and accurately. Effective candidate selection is critical in today’s highly competitive employment market. Traditional resume screening processes are time-consuming and prone to bias, resulting in poor recruiting judgments. This study presents an AI-driven resume screening system based on Natural Language Processing (NLP) to address these difficulties. A proprietary spaCy Named Entity Recognition (NER) library is used to extract key data from resumes, including names, organizations, job titles, skills, experiences, and education. The system employs data preprocessing techniques such as data cleansing, structuring, augmentation, text normalization, tokenization, part-of-speech tagging, stop word removal, and lemmatization/stemming to assure correctness and reliability. Following that, the proprietary NER model computes scores for each application based on the retrieved data, resulting in a ranked list of the best applicants. A comparison study was done to compare the suggested approach to traditional resume screening methods. In terms of accuracy, efficiency, and fairness, the results showed that the NLP-driven resume screening solution surpassed traditional approaches. Furthermore, it lowered the time and effort required for resume screening, allowing recruiters to focus on more important activities like candidate assessments and interviews.

Sarveshwaran R., Karthikeyan S., Cruz M.V., Shreyanth S., Niveditha S., Rajesh P.K. (2024). NLP-Based AI-Driven Resume Screening Solution for Efficient Candidate Selection. In: Yang, XS., Sherratt, S., Dey, N., Joshi, A. (eds) Proceedings of Ninth International Congress on Information and Communication Technology. ICICT 2024 2024. Lecture Notes in Networks and Systems, vol 1012. Springer, Singapore.https://doi.org/10.1007/978-981-97-3556-3_29