Downloads: 0 | Views: 143
Informative Article | Engineering Science | India | Volume 11 Issue 6, June 2022 | Rating: 4.9 / 10
Predictive Incident Management Using Machine Learning
Abstract: One of the significant advances resulting from the ability to predict incidents is the emergence of predictive incident management as the critical capability for proactive monitoring and management of modern complex systems. This survey presents the complete picture of foretelling incident management based on machine learning techniques and the up - to - date advancement in this area. The primary issue is that the way incident management is done on emergency help is reactive and causes people to be delayed, services to be interrupted, and costs to go high. The manual gambling of computerized (hardware/software) failures is almost impossible in complex, highly dynamic systems with a large amount of monitoring data. Machine learning techniques are considered for implementing an automated disaster detection system that can be trained on observing data properties related to system features to forecast incidents before they occur. Applying models like regression and neural networks to supervised learning can help recognize imminent storm symptoms by referencing past antecedents. Another feature of unsupervised techniques such as clustering is that they can be used to identify anomalies witnessed, which gives the earliest indications that there could be emerging problems. Online education goes on to lead to the practicality of predictive software. This review studies the potential of predictive incident management applications in monitoring the IT system, industrial predictive maintenance, healthcare systems, fraud detection, and supply chain risk. The effect, therefore, is improved service quality and stability, diagnostics accuracy, and resource allocation optimization. The following steps are adding new data sources to predict further scenarios, distributed learning at scale, making AIs one step further explaining themselves, and preserving people's privacy while dealing with data. The preference for predictive incident management based on machine learning allows industry - wide organizations to go from reactive to proactive with machine - driven systems management throughout the economy.
Keywords: Predictive analytics, incident management, machine learning, anomaly detection, predictive maintenance, proactive monitoring
Edition: Volume 11 Issue 6, June 2022,
Pages: 1977 - 1980