Bhavesh Ajwani's Projects
This repository contains all my work from Computer Network Lab
Inventory Management System (IMS) is an Android application designed to streamline and simplify inventory management tasks for businesses. With a user-friendly mobile interface, businesses can manage sales, purchases, customer interactions, and get real-time stock valuations at their fingertips.
A machine learning project for SMS message classification, distinguishing between "ham" and "spam" messages using text processing. Enhance communication security with effective spam detection.
A backend infrastructure for collecting, processing, and emitting real-time metrics across services. This scalable solution enhances performance monitoring and integrates seamlessly with existing systems, enabling proactive issue detection and optimization through aggregated metrics and dashboards.
The approach uses adaptive k-means clustering to group tasks based on their similarities, and assigns them to the appropriate Virtual Machines (VMs) based on the availability of resources.
Unleashing Olympic Data Insights, The dashboard is divided into "Medal Tally," "Overall Analysis," "Country Wise Analysis," and "Athlete Wise Analysis," featuring visually captivating data visualizations. Uncover valuable insights into Olympic history, including rankings of participating countries and athletes based on total medals won.
A repository containing code, notes, and examples on Object-Oriented Programming (OOP) concepts, which form the foundation of Low-Level Design (LLD) in software development. Explore key principles such as encapsulation, inheritance, polymorphism, and abstraction.
Researched OSCC detection in histopathological images, involving stain reduction, quality enhancement, deep learning segmentation, and classification to distinguish malignancy. The study compared different segmentation architectures for accuracy.
Applied Q-learning algorithm on gridWorld problem
This railway management system enables users to search for trains, check timings, view ticket fares by compartments, book tickets, store data in a database, and access passenger details for booked tickets.
This project is aimed at implementing a simplified search engine using C++ programming and data structures like maps and tries.
This project implements a simplified search engine using C++ programming and data structures like maps and tries. It is designed to efficiently store and retrieve text documents based on keyword queries.
Graph-based network optimization model to minimize delay and congestion in networks. Unlike conventional algorithms, our model overcomes limitations by considering a broader range of network parameters. Formulated a Deep Reinforcement Learning (DRL) approach with a custom environment, reward function, and agent capable of learning optimal policy.
This repository consists the code of a simple website blocker project implemented in Python. It can be used to block certain websites during work time to reduce distraction thus improving productivity.
The proposed field-specific solution for detecting weeds and crops used series of models and algorithms. Reduced U-Net that works in lesser number of overall parameters is used for Binary semantic segmentation and Weed-Net, (a modified version of U-Net) is used for Multi-class semantic segmentation.
Created a robust weed management system using 800 annotated images, achieving 92% ROI extraction accuracy. Precise weed localization with IoU of 0.85 and multiclass segmentation using a modified U-Net. AI rover with Jetson Nano achieved 0.69 cosine similarity in the field.