Scholars Central

Conference abstractPublished 3 Aug 2022

An ensemble-based scalable approach for intrusion detection using big data framework

Authors: Durga Prasad Mohapatra

National Institute of Technology, India

Ask about this article

Abstract

Select any words in the title or abstract, or press this and choose a sentence.

Statement of the Problem: Big Data analytics plays a vital role in intrusion detection. It provides tools to support structured, unstructured, and semi-structured data for analysis. Also, it offers scalable algorithms for the fast processing of data using machine learning approaches. Most of the AI/ML-based approaches are biased towards the majority class during the learning process. As a result, the minority classes are misclassified. It is a challenging task for researchers of different domains to detect such instances which are commonly known as class imbalanced problems. Research Objective: To develop a scalable, robust intrusion detection model that can handle imbalanced data and avoid biases during the learning and testing process on the big data framework as shown in [Figure 1]. Tools used: Apache Spark, Hadoop, PySpark, etc. Methodology & Theoretical Orientation: In this work, scalable machine learning algorithms i.e. RUSBoost and Decision Tree classifier with revised K-Means approaches for threat classification are implemented using the big data framework. The algorithm has been designed to work even with a relatively small training set and support to classify a large volume of testing data as shown in [Figure 2]. Findings: The result of the proposed approach is discussed. The decision tree is used to detect the majority class and RUSBoost for detecting the minority class in the proposed ensemble approach. The popular intrusion datasets are used for training, validation, and model testing during the experiment. Conclusion: The combination of clustering, boosting, and supervised approach improves the model's robustness to deal with the imbalanced datasets. As per the obtained result, the detection accuracy of the minority classes was drastically enhanced compared to other popular approaches. The classification error is minimized in all the datasets and the confusion matrix and ROC, show the proposed approach's robust detection accuracy.

Ask about this article

Answers come only from this article's title, abstract and keywords, each sentence pointing to where it comes from. Your question is sent to OpenAI to write the answer. Scholars Central does not store it.

Try:

Similar by topic

More from this journal

The newest in Journal of Computer Science & Systems Biology.

All its articles

Conferences in this field

Upcoming conferences in this field.

Follow

New articles from what you follow go into your email digest, once you agree to alert emails.

  • JournalJournal of Computer Science & Systems Biology
  • SubjectComputer Science & IT
  • BylineArticles whose byline prints “Durga Prasad Mohapatra”

For authors and organisers

Publish, present and keep up with your field

Find the right journal for your paper and submit your manuscript.

Photo: Pexels