Anomaly Detection in Data Streams Using Machine Learning and Deep Learning

Data stream mining for movement has emerged as an important area of machine learning because of the huge amount of changing and continuous data coming from diverse sources such as social media, business sensors, and mobile communications. The goal of this anomaly identification in the data streams...

Full description

Saved in:
Bibliographic Details
Main Author: Muhammad Amin Daneshwar
Format: Article
Language:English
Published: International Transactions on Electrical Engineering and Computer Science 2025-01-01
Series:International Transactions on Electrical Engineering and Computer Science
Subjects:
Online Access:https://iteecs.com/index.php/iteecs/article/view/118
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Data stream mining for movement has emerged as an important area of machine learning because of the huge amount of changing and continuous data coming from diverse sources such as social media, business sensors, and mobile communications. The goal of this anomaly identification in the data streams is to find patterns that deviate substantially from the way things usually work. This will be valuable information for making decisions in a large number of areas, including healthcare, management of financial risk, keeping communities safe, and operating the power grid. This research discusses the intractable ways of finding oddities in a stream of data with the corresponding predicaments of always having a new inflow of data, creating information fast, and also dynamics of information changing. We also observe how distinct deep learning and machine learning approaches are being used in different fields to rapidly detect anomalies. Some examples of the way these techniques have been used to discover network intrusions, malware, IoT outliers, healthcare anomalies, and credit card frauds are a demonstration of the techniques work.
ISSN:2583-6471