Identifying Trade Trends in Greek Pottery
Using KModes clustering, silhouette analysis, and fuzzy string matching to unveil the spread and evolution of different pottery styles in ancient Greece.
Posts by Collection
portfolio
Finding Frequent Item Sets in Buried Grave Goods
Using Apriori association rule mining algorithm on Digital Archive for Grave Goods in Later Prehistoric Britain to identify frequently buried together items
Using Classification to Date South African Artefacts
Building a Random Forest Classifier with class weight adjustment to handle skewness to predict the time period to which South African Radio Carbon dated artefacts belong.
Ensemble Learning to find TV Series Ratings Using Text Scripts
Using techniques such as named entity recognition, apriori algorithm and sentiment analysis to convert text script to attributes to predict TV series ratings using Ensemble Learning (Gradient and AdaBoost stacked model).
End-to-End LSTM Pipeline for Song Generation
A Recurrent Neural Network (RNN) based three-stage architecture for end to end music generation.
7 layer CNN Model to Detect Schizophrenia
Using convolutional neural networks and Morlet wavelet transform to analyze EEG data to detect Schizophrenia in adolescents.
Vitals Detection Using Computer Vision
Using computer vision algorithms to detect Heart Rate, SpO2, and Blood Pressure from video feed.
New Neighbourhood Recommender Engine for Movers
K-means clustering based recommendation system to get moving recommendations based on neighborhood data.
Voice Activated AI Virtual Assistant
A voice-assistant service in Python 3.7+ that understands speech, executes commands, and delivers an efficient and personalized user experience.
Optimizing Agriculture Water Usage using AI Driven Sensors
Applying machine learning based WBAN transmission protocols to reduce agriculture water consumption by approximately 38% (in simulation).
Vulnerable-Hackable Django Web Application
Django web application to raise awareness about attack vectors in the Indian cyber space through a hackable interface. Developed during internship with India’s Ministry of Electronics and Information Technology.
publications
PINACC: Pin attack device primarily used as a women’s security provision
Published in The Official Gazette of the Intellectual Property Rights Authority of India, 2020
Patent PINNAC is a security provision that capitalizes on the instinct to defend. Developed over a period of one year (2019-2020), PINACC was filed to be patented in October of 2020. It has since been published in the gazette of The Department of Promotion of Industry and International Trade, Intellectual Property Office, Government of India and the application is currently under consideration.
A Novel Hybrid Clustering Based Transmission Protocol for Wireless Body Area Networks
Published in Computers, Materials & Continua, 2021
Journal Article The research paper proposes a hybrid clustering-based packet propagation protocol to enhance the throughput of wireless sensor networks. By combining machine learning clustering and traditional forwarding functions, the protocol achieves significant improvements in network productivity and throughput values.
Compression using Matrix Folding Algorithm
Published in IEEE 9th International Conference on Reliability, Infocom Technologies and Optimization (Trends and Future Directions) (ICRITO), 2021
Conference Article The research paper presents the Matrix Folding Algorithm, which mathematically reduces storage space for square matrices by 75%, optimizing space utilization in computer programming.
UnClog: A Drone-based Traffic Congestion Mitigation System
Published in Australian Intellectual Property Authority, 2021
Patent UnClog utilizes CNN-based computer vision algorithms and real-time imaging to identify potential waterlogging areas, aiding in early warning to authorities. Now patented with the Australian Intellectual Property Authority
Wireless Sensor Networks (WSNs): Toward an Energy-Efficient Routing Protocol Design
Published in Machine Intelligence and Data Science Application, 2022
Conference Article The primary concern revolving around WSN systems is the energy consumption. The current study employs a cross-layer design approach toward the development of an energy-efficient routing protocol to address this concern.
AI-based security protocols for IoT applications: A critical review
Published in Recent Advances in Computer Science and Communications, 2022
Journal Article An industrywide overview of security provisions is provided by categorizing IoT applications into three broad sectors, namely, healthcare, smart grid, and smart city.
College Recommendation System Using K-Means Algorithm
Published in Futuristic Trends for Sustainable Development and Sustainable Ecosystems, 2022
Conference Article Utilizing massive online data and the foursquare application programming interface, the colleges across the US have been clustered into groups based on the similarity indexes.
Analysis of user inclination in movie posters based on color bias using transfer learning
Published in Emerging Technologies in Data Mining and Information Security, 2022
Conference Article The present article and the underlying research serve to empirically confirm the existence of visual bias that gives rise to the discipline of neuromarketing and put forth a method to quantify such bias using neural style transfer.
talks
Data is Beautiful
Published:
Data is Beautiful webinar celebrated the art of data visualization, empowering participants to unlock the transformative power of visually representing complex datasets and deriving meaningful insights.
Creating a custom locale based college recommendation system using K-means Clustering
Published:
Unveiling the potential of a custom locale-based college recommendation system using K-means clustering, my research showcased the power of personalized recommendations for prospective students, revolutionizing higher education decision-making.
Unsusual Tech Careers
Published:
Embarking on an unusual tech career, founding my own AI/ML lab led to captivating research, inspiring collaborations, and an unforgettable opportunity to share my journey with like-minded individuals.
Towards more energy efficient routing protocols for wireless sensor networks
Published:
This presentation at the Machine Intelligence and Data Science Applications conference highlighted the development of PRRP, an energy-efficient routing protocol for wireless sensor networks, showcasing significant improvements in network performance and energy efficiency.
Using transfer learning to quantify user color bias in movie poster selection for neuromarketing applications
Published:
The results of the study were presented at the IEMIS 2022: 3rd International Conference on Emerging Technologies in Data Mining and Information Security at the Technical Session 1.5 on Advanced Computing where the presented article won the Best Paper Award.
teaching
Teaching experience 1
Undergraduate course, University 1, Department, 2014
This is a description of a teaching experience. You can use markdown like any other post.
Teaching experience 2
Workshop, University 1, Department, 2015
hjhhbk