Vignesh Kumar V Email & Phone Number

Vellore, Tn, In

Vellore, Tamil Nadu, India

Auckland, New Zealand

During my tenure at Signal, I played a pivotal role in optimizing the accuracy and specificity of geo-tagging for intelligence briefs. Working on the CLAVIN-NERD/Stanford CoreNLP model, I focused on enhancing the extraction of location references from unstructured data, thereby improving the precision of security incident tagging. I tackled challenges ranging from atypical text structures to disambiguating between similar locations. Collaborating with a dynamic team, I contributed to refining the model's global inclusivity, ensuring more effective and actionable intelligence for informed decision-making

Chennai, Tamil Nadu, India

I completed the IBM Certified AI Analyst - Associate program, which gave me a solid foundation in the fundamentals of artificial intelligence (AI). I gained hands-on experience with IBM Watson Machine Learning and AutoAI, IBM Watson Knowledge Studio, and IBM Watson Assistant. I built and deployed a regression analysis model with IBM Watson Machine Learning, ingested, converted, enriched, and queried data using NLP techniques, created a chatbot with IBM Watson Assistant, used a basic deep neural network to solve an image recognition problem, and designed and built an intelligent solution for the procurement domain.

Chennai, Tamil Nadu, India

Chennai, Tamil Nadu, India

In this project, we employed a cutting-edge third generation neural network known as eSNN (enhanced Spiking Neural Network) to tackle complex data analysis tasks. Our focus was on enhancing diagnostic capabilities in the medical field. Initially, we meticulously trained the eSNN on a comprehensive Electrocardiogram (ECG) dataset, enabling it to learn intricate patterns and relationships within cardiac data.Taking a step further, we leveraged the power of transfer learning to extend our model's capabilities to a related yet distinct domain. By harnessing the knowledge acquired from ECG data, we seamlessly adapted the eSNN architecture to Electroencephalogram (EEG) datasets. This transfer learning approach allowed our model to efficiently decipher EEG data patterns associated with brain activity.Our work not only showcases the potential of third generation neural networks like eSNN but also underscores the versatility and adaptability of transfer learning. By successfully training our model on both ECG and EEG datasets, we have taken strides towards more accurate medical diagnostics and underscored the potential for shared insights between diverse medical data types. This approach holds promise for broader applications in the realm of healthcare and beyond.

Chennai, Tamil Nadu, India

Explored Spiking Neural Networks (SNNs), the third generation of neural networks for AI applications, inspired by the human brain.Investigated SNNs' efficiency in predicting and classifying data with spatial and temporal information by modeling brain behavior and learning potential.Focused on understanding the functionality of Evolving Spiking Neural Networks (ESNN) as a classifier and differentiating it from existing neural models.Conducted an in-depth examination of the ESNN architecture and fine-tuned various model parameters to enhance prediction accuracies.Applied the optimized ESNN model to diverse datasets and performed a comparative analysis with second-generation neural network models, such as LSTM.Demonstrated substantial improvements in classification accuracy using ESNN, indicating that ESNN and its variants represent the onset of a new era in Neural Networks.