Abstract: A Distributed Denial of Service (DDoS) attack is a cybersecurity weapon aimed at disrupting the operation of services or extorting money from targeted organizations. These attacks can be motivated by politics, religion, competition or profit.

Technically, a DDoS attack is a distributed version of a Denial of Service (DoS) attack whose goal is to disrupt the target's business operations. This type of attack sends a high volume of traffic to overload the normal operation of a service, server, or network interconnection, rendering them unavailable. DoS attacks disrupt service, while Distributed Attacks (DDoS) are carried out on a much larger scale, bringing down entire infrastructures and scalable (cloud) services.

In this work, we have addressed a rather particular DDoS, that relating to amplification. It was therefore a question here not only of talking about it, but also of presenting different forms of attacks, in addition to the related countermeasures for better security of our information systems.

Keywords: Cryptosystem, DDOS, DDoS basé sur l’amplification, information system.

Title: Countermeasures to DDOS attacks based on amplification

Author: KABEYA TSHISEBA Cedric

International Journal of Novel Research in Computer Science and Software Engineering

ISSN 2394-7314

Vol. 9, Issue 3, September 2022 - December 2022

Page No: 29-32

Novelty Journals

Website: www.noveltyjournals.com

Published Date: 29-December-2022

Abstract: Machine learning is a form of artificial intelligence that focuses on creating systems that learn, or improve their performance, based on the data they process. Artificial intelligence is a broad term that refers to systems or machines that simulate a form of human intelligence.

Today, we use machine learning in all areas. When we interact with banks, shop online or use social media, machine learning algorithms come into play to optimize, streamline and secure our experience. Machine learning and the technology around it are developing rapidly, and we are only beginning to glimpse its capabilities. It is exactly within this same framework that we proposed to be able to set up an intelligent system based on artificial intelligence, or precisely on machine learning, to allow schools in difficulty to have one or more orientation advisers on a permanent basis., to be equipped with this service, in order to help secondary school students moving on to the humanities, to be able to choose their options.

Keywords: machine learning, artificial intelligence, orientation advisers, secondary school students.

Title: Implementation of a machine learning-based system for the orientation of secondary school students

Author: KABEYA TSHISEBA Cedric

International Journal of Novel Research in Computer Science and Software Engineering

ISSN 2394-7314

Vol. 9, Issue 3, September 2022 - December 2022

Page No: 24-28

Novelty Journals

Website: www.noveltyjournals.com

Published Date: 20-December-2022

Abstract: Online credit and debit card purchases have increased bank fraud. Using deep learning, researchers analysed cutting-edge fraud-detection algorithms. Most online KNN implementations work with sequential data like time series, audio/speech, and language. KNNs process and store locally, unlike forwarding networks. Time series analysis often uses KNNs. Check the numbers for suspicious purchases. Malaysian customers made 278,309 credit card transactions in October 2022, according to creditcard.csv. According to creditcard.csv, Malaysians used 278,309 credit cards in October 2022. K-nearest neighbour simplifies data categorization. Choose k for accurate scenario classification. k could be any positive integer between 1 and the set's size. When k = 1, a new log with the same category is generated. When k is small relative to the training dataset's features, noise is introduced. A large value of k complicates the calculations so much that a specimen is incorrectly classified into the most frequent category during data training. k-adaptability, KNN's usefulness, and versatility give it many advantages.

Keywords: KNN algorithm, Machine Learning, Bank fraud, Credit/Debit cards.

Title: TRANSACTION FRAUD DETECTOR USING KNN IN DEEP LEARNING

Author: Ahmad Alammar, Yazeed Al Moayed, Nasir Ahmed Algeelani

International Journal of Novel Research in Computer Science and Software Engineering

ISSN 2394-7314

Vol. 9, Issue 3, September 2022 - December 2022

Page No: 16-23

Novelty Journals

Website: www.noveltyjournals.com

Published Date: 26-November-2022

Abstract: Humanoid Robots are enormously complex and that is why the research on humanoids spreads out over all kinds of scientific fields, from mechanics to electronics, from modeling to control and from informatics to biomechanics. This research paper is concerned with the development of a humanoid robot using edge detection technique. In the design process, the model of a humanoid robot was developed first using the simulink library tools in Matlab/Simulink environment. The guassian block convolved the input matrix with the Prewitt kernel and the block outputs two gradient components of the image, the block also performed a thresholding operation on the gradient magnitudes and output a binary image, which is a matrix of Boolean values. During the development of the humanoid robot an object orientated method was employed. Odometry and landmark method was employed in the navigation of the humanoid system. Sample of robot in ordinary form and also in edge detection form was collected and experimented in paint graphics using Portable Network Graphics format. Both of the two samples were scaled to 300X300 in dimension, and also 300X300 pixels with respect to width and height, bit depth as 32 and result showed that the computational size of the ordinary robot was 90.1kb and that of the prewitt edge detection was 16.6kb using Portable Network Graphics. This amounts to 81.58 percentage improvement. The same sample was also collected for canny edge detector and Sobel edge detector and experimented using paint graphics and their computational sizes were recorded as 19.5kb and 29.0kb respectively. This proves that prewitt edge detector is computationally cheaper than Canny as reviewed in (Karl, 2016) which either too much noise observed or too many information omitted.

Keywords: Humanoid robot, Motorized robot, modeling a humanoid robot, Robot obstacle avoidance, Prewitt edge detection, Method of navigation, Robot in canny edge detection, Robot in Roberts edge, Robot in prewitt edge detection form.

Title: ARTIFICIAL INTELLIGENT BASED AUTOMATIC PATH FINDING AND COMPUTATIONAL COMPLEXITY REDUCTION ROBOT USING PREWITT EDGE DETECTION PARADIGM

Author: Chukwuagu M. I, Aneke E.C.

International Journal of Novel Research in Computer Science and Software Engineering

ISSN 2394-7314

Vol. 9, Issue 3, September 2022 - December 2022

Page No: 1-15

Novelty Journals

Website: www.noveltyjournals.com

Published Date: 25-November-2022