Reference no: EM133872060 , Length: word count:1200
Artificial Intelligence for Cyber Security
Learning outcome 1: Develop solutions for data exploration, preprocessing and feature engineering using suitable frameworks (Weka, Matlab or Python).
Learning outcome 2: Design and implement the different learning models available in neural networks, to develop solutions to neural network-based problems.
Learning outcome 3: Analyze and compare the models based on their performance metrics and discuss their advantages and limitations.
Task 1: Data Exploration, Loading and Preprocessing
You will be provided on Learning Zone with the networkintrusiondetectiondataset UNSW- NB15 dataset:
Loading and Preprocessing the dataset.
Handling Missing Values using appropriate strategies, Report your imputation decisions in a brief summary.
Task 2: Correlation Analysis and Feature Selection
Correlation Analysis: Visualize correlations between features.
Feature Selection: Apply appropriate feature selection, such as correlation-based feature selection and wrapper method to select the most valuable features.
Task 3: Model Training and Evaluation
Analyse a supervised machine learning approach, such as tree-based classifier, and rule-based classification approach PART. Train the model with a k-fold cross- validation technique and/or full training set with reporting experimental results such as classification accuracy, TPR, FPR and visualization of trees. Get online assignment help in the USA!
Analyse the unsupervised learning approach, i.e., clustering (simple k- means), train the model and visualize the output with a clustering plot.
Task 4: Model Comparison
Compare supervised and unsupervised approaches individually and explain which performed better.
Compare the performance of the supervised and unsupervised learning approach in terms of Confusion metrics (TPR, FPR), and overall accuracy.
A short discussion explaining the comparison with a comparison table.
Task 5: Report and Code Submission
Summarize your findings and submit your work. The report should be organised as an introduction, all task descriptions and a conclusion.
In the Learning Zone you will find the link to submit the following:
Report: A report covering all parts of the coursework, including code submission (Weka, MATLAB, or Python Code), which is clearly labelled, with sections corresponding to each task.