This coursework (CRWK) must be attempted in the groups of 4 or 5 students. This coursework is divided into two sections: (1) Big Data analytics on a real case study and (2) group presentation. All the group members must attend the presentation. Presentation would be online through Microsoft Teams. If you do not turn up in the presentation date with the video call, you will fail the module.
CN7031 Big Data Analytics Assignment-East London University UK
Overall mark for CRWK comes from two main activities as follows:
1.Big Data Analytics report (around 3,000 words, with a tolerance of ± 10%) in HTML format (60%)
2.Presentation (40%)
Marking Scheme:
Big Data Analytics using Spark SQL
(6) Providing 2 queries using Spark SQL.
(14) Developing advanced SQL statements. Refer to:
(10) Visualizing the outcomes of queries into the graphical and
textual format, and be able to interpret them.
Big Data Analytics using Py Spark
(45) Analyzing the dataset through 3 statistical analytics methods including advanced descriptive statistics, correlation, hypothesis testing, density estimation, etc.
(15) Designing one classifier, then evaluate and visualize the accuracy/performance. Applying a multi-class classifier is considered for full mark.
Documentation
(10) Write down a well-organized report for a programming and analytics project.
IMPORTANT:you must use CRWK template in the HTML format, otherwise it will be counted as plagiarism and your group mark would be zero. Please refer to the “THE FORMAT OF FINAL SUBMISSION” section.
(1) Understanding Dataset:
This data set was originally created by the University of New Brunswick for analyzing DDoS data. You can find the full dataset and its description here. The dataset itself was based on logs of the university’s servers, which found various DoS attacks throughout the publicly available period to generate totally 80 attributes with 6.40 GB size. We will use about 2.6 GB of the data to process it with the restricted PCs to 4GB RAM. Download it from here. When writing machine learning or statistical analysis for this data, note that the Label column is arguably the most important portion of data, as it determines if the packets sent are malicious or not.
CN7031 Big Data Analytics Assignment-East London University UK
a) The features are described in the “IDS2018_Features.xlsx” file in Moodle page.
b) The labels are as follows:
• “Label”: normal traffic
• “Benign”: susceptible to DoS attack
c) In this coursework, we use more than 8.2-million records with the size of 2.6 GB. As a big data specialist, firstly, we should read and understand the features, then apply modeling techniques. If you want to see a few records of this dataset, you can either use [1] Hadoop HDFS and Hive, [2] Spark SQL or [3] RDD for printing a few records for your understanding.
(2) Big Data Query & Analysis using Spark SQL
This task is using Spark SQL for converting big sized raw data into useful information. Each member of a group should implement 2 complex SQL queries (refer to the marking scheme). Apply appropriate visualization tools to present your findings numerically and graphically. Interpret shortly your findings.
• What do you need to put in the HTML report per student?
- At least two Spark SQL queries.
- A short explanation of the queries.
- The working solution, i.e., plot or table.
• Tip: The mark for this section depends on the level of your queries complexity, for instance using the simple select query is not supposed for a full mark.
(3)Advanced Analytics using Py Spark
In this section, you will conduct advanced analytics using PySpark.
3.1. Analyze and Interpret Big Data using PySpark
Every member of a group should analyze data through 3 analytical methods (e.g.,advanced descriptive statistics, correlation, hypothesis testing, density estimation, etc.). You need to present your work numerically and graphically. Apply tool tip text, legend, title, X-Y labels etc. accordingly.
Note: we need a working solution without system or logical error for the good/full mark.
3.2. Design and Build a Machine Learning (ML) technique
Every member of a group should go and apply one ML technique. You can apply one the following approaches: Classification, Regression, Clustering, Dimensionality Reduction, Feature Extraction, Frequent Pattern
mining or Optimization. Explain and evaluate your model and its results into the numerical and/or graphical representations.
Note: If you are 4 students in a group, you should develop 4 different models. If you have a similar model, the mark would be zero.
(4) Documentation
Your final report must follow the “The format of final submission” section. Your work must demonstrate appropriate understanding of building a user friendly, efficient and comprehensive analytics report for a big data project to help move users (readers) around to find the relevant contents.
THE FORMAT OF FINAL SUBMISSION
1.You can use either Google Co lab or Ubuntu V M Ware for this CRWK.
2.You have to convert the source code (*.ipynb) to HTML. Watch the video in the Moodle about “how to submit the report in HTML format”.
3.Upload ONLY one single HTML file per group into Turnitin in Moodle. One member of each group must submit the work, NOT all members. The name of the file must be in the format of “Your-Group-ID_CN7031”, such as Group 200_CN7031 .html if you are belonging to the group 200.
4.The submission link will be available from week 10, and you are free to amend your submitted file several times before submission deadline. Your last submission will be saved in the Moodle database for marking.
CN7031 Big Data Analytics Assignment-East London University UK
PLAGIARISM
If there are copied PySpark codes from somewhere or someone else, all the group members will get zero, and should attend the “breach of regulation” committee for further explanations and the probable additional penalties.