Analysing the Influence of Activation Functions in CNN models for Effective Malware Classification
- Title
- Analysing the Influence of Activation Functions in CNN models for Effective Malware Classification
- Creator
- Kasliwal S.; Vinay M.; Jayapriya J.; Deepa S.; Mahalakshmi J.
- Description
- With the advancement of information technology, malware has become a persistent cyber security concern that targets computer systems, smart devices, and wide networks. Due to flaws in performance accuracy, analysis type, and malware classification methodologies that miss unsuspected malware attacks, malware classification has thus always been a significant concern and a challenging subject. Using the Malimg dataset, which has 9349 samples from 25 different families, this study classifies malware using a deep learning algorithm called a convolution neural network and evaluating the accuracy using a number of activation functions in this study. The proposed CNN model for malware classification achieves a high accuracy rate without the need for complex feature engineering. The model achieved the highest accuracy of 96.93% when using the Rectified Linear Unit (ReLU) activation functions whereas Leaky Relu gives accuracy of 96.76%, Pre relu gives 96.36%, ELU gives 95.72% and tanh gives accuracy of 95.58%. 2024 IEEE.
- Source
- 2nd IEEE International Conference on Data Science and Network Security, ICDSNS 2024
- Date
- 2024-01-01
- Publisher
- Institute of Electrical and Electronics Engineers Inc.
- Subject
- activation function; cnn; deep learning; malware classification
- Coverage
- Kasliwal S., Christ Univesity, Department of Computer Science, Bangalore, India; Vinay M., Christ Univesity, Department of Computer Science, Bangalore, India; Jayapriya J., Christ Univesity, Department of Computer Science, Bangalore, India; Deepa S., Christ Univesity, Department of Computer Science, Bangalore, India; Mahalakshmi J., Christ Univesity, Department of Computer Science, Bangalore, India
- Rights
- Restricted Access
- Relation
- ISBN: 979-835037311-0
- Format
- Online
- Language
- English
- Type
- Conference paper
Collection
Citation
Kasliwal S.; Vinay M.; Jayapriya J.; Deepa S.; Mahalakshmi J., “Analysing the Influence of Activation Functions in CNN models for Effective Malware Classification,” CHRIST (Deemed To Be University) Institutional Repository, accessed February 25, 2025, https://archives.christuniversity.in/items/show/19134.