Open Access Original Research Article

Machine learning Algorithm of Intrusion Detection System

Rozin Majeed Abdullah, Adnan Mohsin Abdulazeez, Adel Al-Zebari

Asian Journal of Research in Computer Science, Volume 9, Issue 3, Page 1-12
DOI: 10.9734/ajrcos/2021/v9i330221

Web of thing (WoT) is a gifted answer for interface and access each gadget through the web. Consistently the gadget includes increments with huge variety fit as a fiddle, size, use and intricacy. In this paper Since WoT drives the world and changes individuals' lives with its wide scope of administrations and applications. In any case, WoT offers various types of assistance through applications, it faces serious security issues and powerless against assaults, for example, sinkhole assault, overhang dropping, forswearing of administration assaults. So on, the Interruption recognition framework is utilized to recognize such assaults when the organization's security is penetrated. Given a scale extension of Web of Things for a practical asset the executives in brilliant urban communities, a legitimate plan of an interruption recognition framework IDS is basic to protect the future organization framework from interlopers. With the development of associated things, the most broadly utilized brought together cloud-based IDS regularly suers from high inertness and organization overhead, subsequently coming about in lethargy to assaults and moderate recognition of pernicious clients.

Open Access Original Research Article

Predicting Weather Forecasting State Based on Data Mining Classification Algorithms

Fairoz Q. Kareem, Adnan Mohsin Abdulazeez, Dathar A. Hasan

Asian Journal of Research in Computer Science, Volume 9, Issue 3, Page 13-24
DOI: 10.9734/ajrcos/2021/v9i330222

Weather forecasting is the process of predicting the status of the atmosphere for certain regions or locations by utilizing recent technology. Thousands of years ago, humans tried to foretell the weather state in some civilizations by studying the science of stars and astronomy. Realizing the weather conditions has a direct impact on many fields, such as commercial, agricultural, airlines, etc. With the recent development in technology, especially in the DM and machine learning techniques, many researchers proposed weather forecasting prediction systems based on data mining classification techniques. In this paper, we utilized neural networks, Naïve Bayes, random forest, and K-nearest neighbor algorithms to build weather forecasting prediction models. These models classify the unseen data instances to multiple class rain, fog, partly-cloudy day, clear-day and cloudy. These model performance for each algorithm has been trained and tested using synoptic data from the Kaggle website. This dataset contains (1796) instances and (8) attributes in our possession. Comparing with other algorithms, the Random forest algorithm achieved the best performance accuracy of 89%. These results indicate the ability of data mining classification algorithms to present optimal tools to predict weather forecasting.

Open Access Original Research Article

On the leaning of the widespread adaptation of web services such as social networking sites (like Twitter, Facebook, LinkedIn, YouTube, WhatsApp, Instagram, Pinterest, etc.) and E-mail have become regular work. We approach these sites to gather or share information worldwide in the form of messages (like tweets, posts, blogs, etc.) and also in other formats such as pictures, audio, and video. In the modern era of Technology where the audience is widely connected with e-platform, these social networking sites are also used to organize e-campaign to favor or counteraction in different contention such as political review, social issue, environmental dispute, worldwide controversy, trolling etc. using the method of Folksonomy [1]. We are participating in such trolling, controversy, and campaign or expedition by using posting a message, tweet, micro-blog, etc. In particular, to join all we are doing is post a tweet or micro-blog that has the precise word or phrase because it appears within the Trends list, like a hashtag. But the trending keywords changed in the short-term and any hashtag gets popularity worldwide shortly. We demonstrate the custom-URL to join e-campaign which is wrapped in shortened-URL for easy to understand and gets excessive results to trend any Tag or Hashtag in a span of time. We improve the results for the community, groups and as well as for the individual audience to gets the best consequence for trending keywords.

Open Access Original Research Article

Leukemia Diagnosis using Machine Learning Classifiers Based on Correlation Attribute Eval Feature Selection

Revella E. A. Armya, Adnan Mohsin Abdulazeez, Amira Bibo Sallow, Diyar Qader Zeebaree

Asian Journal of Research in Computer Science, Volume 9, Issue 3, Page 52-65
DOI: 10.9734/ajrcos/2021/v9i330225

Leukemia refers to a disease that affects the white blood cells (WBC) in the bone marrow and/or blood. Blood cell disorders are often detected in advanced stages as the number of cancer cells is much higher than the number of normal blood cells. Identifying malignant cells is critical for diagnosing leukemia and determining its progression. This paper used machine learning with classifiers to detect leukemia types as a result, it can save both patients and physicians time and money. The primary objective of this paper is to determine the most effective methods for leukemia detection. The WEKA application was used to evaluate and analyze five classifiers (J48, KNN, SVM, Random Forest, and Naïve Bayes classifiers). The results were respectively as follows: 83.33%, 87.5%, 95.83%, 88.88%, and 98.61%, with the Naïve Bayes classifier achieving the highest accuracy; however, accuracy varies according to the shape and size of the sample and the algorithm used to classify the leukemia types.

Open Access Review Article

Image Authentication Based on Watermarking Approach: Review

Basna Mohammed Salih Hasan, Siddeeq Y. Ameen, Omer Mohammed Salih Hasan

Asian Journal of Research in Computer Science, Volume 9, Issue 3, Page 34-51
DOI: 10.9734/ajrcos/2021/v9i330224

Digital image authentication techniques have recently gained a lot of attention due to their importance to a large number of military and medical applications, banks, and institutions, which require a high level of security. Generally, digital images are transmitted over insecure media, such as the Internet and computer networks of various kinds. The Internet has become one of the basic pillars of life and a solution to many of the problems left by the coronavirus. As a result, images must be protected from attempts to alter their content that might affect important decision-making. An image authentication (IA) system is a solution to this difficult problem. In the previous literature, several methods have been proposed to protect the authenticity of an image. Digital image watermark is a strategy to ensure the reliability, resilience, intellectual property, and validity of multimedia documents. Digital media, such as images, audio, and video, can hide content. Watermarking of a digital image is a mechanism by which the watermark is embedded in multimedia and the image of the watermark is retrieved or identified in a multimedia entity. This paper reviews IA techniques, watermark embedding techniques, tamper detection methods and discusses the performance of the techniques, the pros and cons of each technique, and the proposed methods for improving the performance of watermark techniques.