Indika Sampath Sirimanna

Blockchain-based Secure, Reliable, and Distributed Voting System for Decision Making in Government Policies and Projects

Abstract — A large part of society today does not trust decision making in government policies and projects. It is time to move into new technology such as decision making on government policies and projects using the latest technology like blockchain and advanced cryptography. Building of blockchain-based secure, reliable, and distributed voting system for decision making in government policies and projects has been a challenge for a long time.

Distributed ledger technologies are exciting technology advancement in the information technology world. Blockchain-based technologies offer an infinite range of applications that benefit from sharing economies of scale. The purpose of this research paper is to evaluate the use of blockchain as the implementation of distributed electronic voting systems for decision making on government policies and projects.

More generally, this article assesses the potential of distributed ledger technologies in improves security by implementing an election process and blockchain-based application. This technology will be benefited to Grama sewa officers who will give correct vote. Because country like Sri Lanka we have 14020 Grama sewa regions. Each member get a chance to cast his vote. One vote for each project, it can be changed before the due date and only the last one can be counted.

The grama sewa officer who cast his vote cannot be tracked, but the officer can cast a unique vote. After a while, the officer may join a fund or a fine. Voting on government policies or projects is mandatory and officer cannot prevent it. This will help to get the best option for government policies or projects.

INTERNET OF THINGS MEETS BRAIN-COMPUTER INTERFACE

Abstract — Brain exploration was a thing of the past and is still a huge field of research. Although a large number of researchers are looking for answers on brain function, it is not yet fully defined. The human brain has a large number of neurons in the brain, and everything in the human body is responsible for them and their function. The brain is the most complex organ in the human body, which controls all of the body's actions and reactions. This is done by receiving various stimuli through the nervous system.

Messages are repositioned through the nervous system in the form of electrical pulses that arise at the moment of stimulation. As soon as a stimulus exceeds the threshold, stimulus-sensing neurons generate information about the frequency of the stimulus and the action potential. When action potentials are generated, they pass through the neural network and eventually reach the brain. Different areas of the brain respond to the type and location of the stimulus.

Today, people can measure their brain waves and patterns outside of medical laboratories. Furthermore, in addition to analysing brain signals, these brain signals can be used as a means of controlling everyday electronic devices, also known as the brain-computer interface. This research focuses on using brain waves for controlling electronic devices, controlling the state of Mimosa Pudica (Shameplant), and reading alphabetical characters on mind by using eye blinking, specifically designed for disabled people with tetraparesis.

Predictive Analytics for Organizational Meal Consumption Using Nearest Neighbor Machine Learning Algorithms

Abstract — This research presents a novel approach to predicting organizational meal consumption patterns using nearest neighbor machine learning algorithms. In large organizations, accurate forecasting of meal requirements is crucial for budget planning, reducing food waste, and ensuring optimal resource allocation. Our study implements a k-nearest neighbor (k-NN) algorithm to analyze historical meal consumption data and predict future requirements with high accuracy.

The methodology involves preprocessing of daily meal consumption data collected over three years from a mid-sized organization with 500+ employees. We engineered features including seasonality indices, day-of-week effects, holiday proximity indicators, and special event markers. The nearest neighbor algorithm was trained on this multidimensional feature space to identify patterns and relationships between organizational activities and meal consumption volumes.

Our model achieved 92.7% accuracy in predicting daily meal requirements for a test period of three months. More impressively, when validated against a full year of consumption data, the model maintained 89.3% accuracy, demonstrating its robustness for long-term forecasting. The algorithm successfully captured seasonal variations, employee attendance fluctuations, and the impact of organizational events on cafeteria usage.

This predictive system enables organizations to optimize food procurement, staffing, and preparation up to one year in advance. Implementation in the test organization resulted in a 27% reduction in food waste and 18% cost savings in the first six months of deployment. The methodology developed can be adapted to various organizational settings with minimal customization requirements, offering a practical solution to the challenge of meal demand forecasting in institutional settings.