Abstract – In the rapidly evolving landscape of software development, agility has emerged as a cornerstone principle in navigating the complexities of modern projects. Agile Software Development is a paradigm that prioritizes adaptability, collaboration, and iterative progress, challenging traditional, rigid development approaches. This paper examines and analyzes three prominent software development methodologies: Agile Software Development, Lean Development, and Dynamic Systems Development, with a focus on key parameters such as project requirements, user involvement, development team dynamics, type of project, and associated risks. Each methodology is introduced, providing insights into their fundamental principles and methodologies. The analysis systematically assesses the suitability and effectiveness of these approaches in various project scenarios, shedding light on their unique strengths and weaknesses. By scrutinizing how each methodology addresses specific project requirements, engages users, forms development teams, caters to project types, and mitigates risks, this paper aims to provide valuable insights to assist organizations in making informed decisions regarding the selection and implementation of the most appropriate development methodology for their projects.

Keywords – Software Development Models, Agile Process, High and Light weight models

DOI: 10.5281/zenodo.12567911

⇩  PDF

Abstract – This conference paper delves into the critical issue of student retention in higher education institutions and proposes a comprehensive set of strategies to address this challenge. As the landscape of education evolves, the need for effective retention measures becomes increasingly paramount. This paper draws on a combination of research findings, case studies, and best practices to provide a holistic framework for improving student retention rates. From academic support to fostering a sense of belonging, the proposed strategies aim to create an environment that encourages students to persist and succeed.

Keywords – Student retention, higher education, academic support, engagement, sense of belonging.

DOI: 10.5281/zenodo.12571672

⇩  PDF

Abstract – It is a high concern to secure huge amount of imaging data stored over the cloud servers. The Visual Cryptography (VC) is a widely used approach to encrypt these imaging data. VC is a powerful technique in which a secret image can be divided into two or more shares and the decryption can be done using human visual system. The VC may understand as crypto sharing approach for embedding true crypto image information to the transparency ciphers. VC has wide range of applications like in biometrics, print online banking, cloud computing, internet voting, etc. In VC a secret image is hidden into two or more shares which on superimposing will recover the hidden image. There are many algorithms designed for VC to secure the images. A related survey has been done in this paper on various visual cryptography schemes based on the number of secrets, pixel expansion, type of share generated, image format, and number of secret images. Paper also presents a detailed review about various visual cryptography color images for cloud storage. 

Keywords – Visual Cryptography, Image, Security, Halftoning, Multi-Share, Encryption, Cloud.

DOI: 10.5281/zenodo.12571246

⇩  PDF

Abstract – Agriculture plays an important role in Indian economy. But now-a-days, agriculture in India is undergoing a structural change leading to a crisis situation. The only remedy to the crisis is to do all that is possible to make agriculture a profitable enterprise and attract the farmers to continue the crop production activities. As an effort towards this direction, this research paper would help the farmers in making appropriate decisions regarding the cultivations with the help of machine learning. This paper focuses on predicting the appropriate crop based on the climatic situations and the yield of the crop based on the historic data by using Random Forest machine learning algorithms. this paper proposes an idea to predict the crop and yield of the crop based on the climatic conditions and historic data related to the crop. The farmer will check the production of the crop as per the acre, before cultivating onto the field. The quantity of grains required by the population in a given year is heavily influenced by population growth and weather changes.

Keywords – Machine learning, cultivation, decisions, web Application

DOI: 10.5281/zenodo.12578772

⇩  PDF

Abstract – With the increasing prominence of Wireless Sensor Networks (WSNs), addressing fault diagnosis has become a pivotal research concern. The emergence of faulty nodes, often stemming from energy depletion, poses significant challenges to the network’s communication reliability and performance. This paper introduces the Energy Efficient Delay Sensitive (EEDS) algorithm as a solution to enhance both energy efficiency and delay management in the presence of faulty nodes. The proposed EEDS algorithm leverages Particle Swarm Optimization (PSO), a well-established optimization technique, to determine an optimised route between source and destination nodes. The algorithm considers the residual energy of nodes as a key factor in initiating communication, ensuring efficient utilisation of available resources. Additionally, the EEDS method employs the Ad Hoc On-Demand Multipath Distance Vector (AOMDV) routing protocol to establish a multipath route, enhancing network robustness. This paper comprehensively details the working of the PSO process, the network model, energy model, fault model, and presents a flowchart along with the algorithmic steps of the EEDS method. The proposed approach not only addresses the challenges associated with faulty nodes but also contributes to minimising energy consumption, thus extending the overall lifetime of the network. The effectiveness of the EEDS algorithm is validated through simulations, demonstrating its potential to significantly improve the fault-tolerant capabilities of WSNs in real-world scenarios. 

DOI: 10.5281/zenodo.12579224

⇩  PDF

Abstract — The evolution of medical imaging technology has sparked revolutionary progress in neuroimaging, particularly in the domain of brain tumor diagnostics. This survey paper navigates through cutting-edge methodologies in brain tumor classification, centering on the transformative impact of deep convolutional neural networks (CNNs). The integration of deep learning techniques, specifically CNNs, has reshaped the land- scape of brain tumor classification by automating the extraction of intricate features from medical imaging data, notably magnetic resonance imaging (MRI) scans. The review critically evaluates key studies that leverage CNN architectures for brain tumor classification, emphasizing diverse datasets, model architectures, and evaluation metrics. Furthermore, the review explores the integration of CNNs with traditional architectures, underscoring the innovative approaches to enhance classification accuracy. As a synthesis of contemporary research, this survey paper aims to furnish a comprehensive understanding of the current landscape of brain tumor classification using deep convolutional neural networks. By critically assessing methodologies, achievements, and challenges, it endeavors to guide future research directions, aspiring to refine diagnostic accuracy, optimize model performance, and ultimately advance personalized treatment strategies for individuals grappling with brain tumours.

Keywords — Brain tumor, Magnetic resonance imaging (MRI), Deep learning, convolutional neural networks (CNN)

DOI: 10.5281/zenodo.12593345

⇩  PDF