Security of edge devices using hybrid machine learning approach with NSL-KDD dataset
Budget: ₹12,500 – ₹37,500 INR
Edge computing is much vulnerable to various types of security attacks. Various approaches for detecting security threats in edge computing have been presented. Machine learning has been proposed by various researchers for attack detection. The schemes which have already been proposed have various phases of data preprocessing, feature extraction, and classification. It is analyzed that machine learning algorithms like SVM, KNN, DT, etc. give low accuracy for vulnerability detection. A novel machine learning algorithm is required which increases accuracy for vulnerability detection in edge computing.
Note- [ Need new hybrid ML algorithm, not an existing one with GitHub available code.]
The main objective of this project is to secure edge devices using a hybrid machine learning algorithm that gives better accuracy than the previously proposed method. For example using a hybrid of random forest, decision forest, KNN, and SVM was used prior which gives an accuracy of 96.28. So, I need a mixture of algorithms that performs better than 96.28%. I also need normal classifiers for comparisons purpose. Also, I have added a sample idea that specifies the flow which you can modify.
Note- [ Need new hybrid ML algorithm, not an existing one with GitHub available code.]
The main objective of this project is to secure edge devices using a hybrid machine learning algorithm that gives better accuracy than the previously proposed method. For example using a hybrid of random forest, decision forest, KNN, and SVM was used prior which gives an accuracy of 96.28. So, I need a mixture of algorithms that performs better than 96.28%. I also need normal classifiers for comparisons purpose. Also, I have added a sample idea that specifies the flow which you can modify.