Journal of Computational Learning Strategies & Practices

Sequential Johnson’s APSP Algorithm on GPU

Anila Batool — 2021-12-12

A quite ordinary issue while processing graphs is to find the shortest distance from one node to all the other nodes. It is called all-pairs shortest path. The applications of finding shortest paths are several including Digital Mapping, Social Networking, Telephone Networks, IP Routing, Fighting Agenda, Robotic Path and many more. Although a lot of work has already been carried out in this aspect but it has been observed that it is quite difficult to exquisitely process graphs which contain a very larger number of nodes. This paper aims to provide a pertinent solution to this problem by proposing three different versions of Johnson’s shortest path solution in parallel architecture over Graphic Processing Unit which resolves APSP problem. As compared to processing extensively large graphs on CPU, proposed architecture will provide a 4.5 time efficient solution for APSP problem.

Prioritizing Effectiveness of Algorithms of Association Rule Mining

Nazish Ghafoor — 2021-11-16

From the last decade, clouds have become the most popular platform for data storage. In the current age, people and machines are engaged in transferring data on clouds. The devices like cameras, computers, mobile, and CCTV are being used to transmit data collectively on the clouds in each second. Such a huge collection of data on clouds is known as Big data. Data mining is a process of extracting useful information from a set of huge data. The different techniques like Association Rule Mining, Classification, and Clustering are some of the well-known techniques, which can be used for data mining purposes. Association Rule Mining is a process of mining associations and correlations among the items in a large data set. Some traditional algorithms like Apriori, FP-growth, and Eclat are being used for association rule mining, but no one provides an optimal solution. In our study, we ascertained the working algorithms, evaluated their performance, and finally ranked them based on their efficiency. We adopted a quantitative approach in our research. We framed queries, pinpointed pertinent work, gauged quality, summarized the evidence, and finally interpreted our findings.

Reviewing Matrix Muliplication

Neelam Amien — 2021-07-13

Algorithm written using different methods can still give same result. One of the ways to determine whether a solution is optimal or not is to determine how much time does it take to solve the specific problem. The problem that is targeted in this paper is Matrix multiplication that is widely used in many scientific computations. Different solutions of this problem are evaluated in this paper and all of them are compared on the basis of their time complexity. After comparing five most known algorithms for matrix multiplication we concluded that coppersmith-winograd algorithm is fastest in terms of time.

A Survey of Trilogy Shortest Path Algorithms

Shaista Sarwar — 2021-06-20

Shortest path problem is one of the most classical problem in graph theory, aiming to discover the shortest path between two nodes in a graph. In this problem, we have to find the minimum-cost tracks or shortest paths between the starting node and final destination in a given graph.Thiswork gives a brief introduction of the most famous algorithms of the shortest path problem i.e. Dijkstra’s, Bellman-Ford, and Floyd Warshall algorithm. A comparative analysis of these algorithms is performed based on their advantages, disadvantages, and efficiency, and application areas.

Genetic Algorithm Inspired Task Scheduling Optimization in Cloud Environment

RAVEEL ABDULLAH — 2021-05-29

With the advancements in various fields of technology, complex and data-oriented problems require supercomputers for proper computation. Massive data that is gathered from various fields of sciences and engineering and it is increasing exponentially with every passing day. There is a dire need of economical solution for efficient processing of data. This is where cloud computing comes into picuture. Cloud computing plays a major role in providing services to individual people and companies with added benefits such as elastic computation, pay on the go, scalable and high-performance computation solutions. The performance factor in cloud environments is highly reliant on task scheduling. Load balancing is simply the distribution of incoming load of users’ request onto the available computing machines. This paper aims to cover a brief overview of Cloud computing and Genetic algorithm and implementation of genetic algorithm for task scheduling pupose. The aim of this paper is to perform a comparative analysis of various task scheduling algorithms that are already in practice and are used in cloud computing. The paper also covers improved scheduling techniques that are inspired from Genetic Algorithm. A systematic comparative analysis of different task scheduling algorithms is also included. Lastly some concluding remarks and future work inspiration is described.