assignment of data science practice... in assignment both coding and report have to complete
ICT707 Data Science Practice Task 3 Semester 2, 2019 ICT707 Data Science Practice Assignment 3 Page 2 of 6 Assessment and Submission Details Marks: 40% of the Total Assessment for the Course Due Date: 11:59pm Friday, Exam Week 1 Submit your assignment to Blackboard Task 3. Please follow the submission instructions in Blackboard. The assignment will be marked out of a total of 100 marks and forms 40% of the total assessment for the course. ALL assignments will be checked for plagiarism by SafeAssign system provided by Blackboard automatically. Refer to your Course Outline or the Course Web Site for a copy of the “Student Misconduct, Plagiarism and Collusion” guidelines. Late submission will be penalised according to the policy in the course outline. Please note Saturday and Sunday are included in the count of days late. Requests for an extension to an assignment MUST be made to the course coordinator prior to the date of submission and requests made on the day of submission or after the submission date will only be considered in exceptional circumstances. Assignment submission extensions will only be made using the official University guidelines. ICT707 Data Science Practice Assignment 3 Page 3 of 6 Assignment Task This assignment consists of two deliverables, being: • One code implementation (50%). The code file in Jupyter Notebook format and the relevant date set files should be contained within a folder named: Task 3-Your Name- Student_Number, the folder is then to be zipped and uploaded to blackboard. • A report (50%). The report must be uploaded as a separate file. Part I - PySpark source code (50%) Important Note: For code reproduction, your code must be self-contained. That is, it should not require other libraries besides PySpark environment we have used in the workshops. The data files are packaged properly with your code file. In this component, we need to utilise Python 3 and PySpark to complete the following data analysis tasks: 1. Exploratory data analysis 2. Recommendation engine 3. Classification 4. Clustering You need to choose a dataset from Kaggle (https://www.kaggle.com/datasets) to complete these tasks. Remember to include the data set file in you source code submission. Note: In your notebook, please use Heading 1 Markdown cell to separate each sub task. Task I.1: Exploratory data analysis This subtask requires you to explore your dataset by • telling its number of rows and columns, • doing the data cleaning (missing values or duplicated records) if necessary • selecting 3 columns, and drawing 1 plot (e.g. bar chart, histogram, boxplot, etc.) for each to summarise it Task I.2: Recommendation engine This subtask requires you to implement a recommender system on Collaborative filtering with Alternative Least Squares Algorithm. You need to include • Model training and predictions • Model evaluation using MSE Task I.3: Classification This subtask requires you to implement a classification system on Logistic regression with LogisticRegressionWithLBFGS class. You need to include https://www.kaggle.com/datasets ICT707 Data Science Practice Assignment 3 Page 4 of 6 • Logistic Regression model training • Model evaluation Task I.4: Clustering This subtask requires you to implement a clustering system on K-means. You need to include • Model training • Model evaluation Part II –Report (50%) You are required to write a report to explain your design and implementation of the machine learning parts in your code, including the following topics: • Introduction/summary/explanation to the ML algorithm/concepts • The learning settings, such as how to prepare training/testing set, what are the key parameters and how you set them up • Comments/evaluation for the models learnt Your report should use the following template: Table of Contents 1.0 Introduction Explain the data set you’ve chosen, including its source URL. Demonstrate your exploratory data analysis in this section. 2.0 Machine learning implementation 2.1 Collaborative filtering 2.3 Logistic regression 2.4 K-Means 3.0 Conclusion References The marking rubrics are viewable on the blackboard. Report Format Your report should be about 1000 words, but no more than 1500 words. The report MUST be formatted using the following guidelines: • Title Page – Must not contain headers, footers, or page numbering. Include your name as the report’s author. • Header – Report title • Footer – your name and the page number ICT707 Data Science Practice Assignment 3 Page 5 of 6 • Paragraph text – 12 point Calibri single line spacing • Headings – Arial in an appropriate type size • Margins – 2.5cm on all margins • Page numbering • Introduction and onwards to use conventional numerals (1, 2, 3, 4) starting at page 1 from the introduction. • The report is to be created as a single Microsoft Word document (version 2007 or later). No other format is acceptable and doing so will result in the deduction of marks. Please follow the conventions detailed in: Summers, J. & Smith, B., 2014, Communication Skills Handbook, 4th Ed, Wiley, Australia. Referencing The report is to include (at least 5) appropriate references and these references should follow the Harvard method of referencing. Note that ALL references should be from journal articles, conference papers, technical papers or a recognized expert in the field. DO NOT use Wikipedia as a reference. The use of unqualified references will result in the deduction of marks. Assignment Return and Release of Grades Assignment grades will be available on the blackboard in two weeks after the submission. Details of marking will also be accessible via online rubrics on the blackboard. Where an assignment is undergoing investigation for alleged plagiarism or collusion the grade for the assignment and the assignment will be withheld until the investigation has concluded. Assignment Advice This assignment will take several weeks to complete and will require a good understanding of machine learning and PySpark for successful completion. It is imperative that students take heed of the following points in relation to doing this assignment: 1. Ensure that you clearly understand the requirements for the assignment – what must be done and what are the deliverables. 2. If you do not understand any of the assignment requirements – Please ASK your tutor. 3. Each time you work on any aspect of the assignment reread the assignment requirements to ensure that what is required is clearly understood. 4. We have practiced nearly all coding tasks in DataCamp before. If you have any difficulty, redoing the practices in DataCamp is recommended. 5. Prior to submitting your code, you should ensure not only that it executes as required, but also looks professional. It is expected that you adhere to python standards for naming and ICT707 Data Science Practice Assignment 3 Page 6 of 6 indenting. All methods should be adequately documented such that another programmer examining your code will readily know what the code is doing. Plagiarism and Collusion Advice 1. All work must be submitted through SafeAssign. 2. SafeAssign will pick up any similarities between work online as well as work from other students (in this semester and previous) 3. Please make sure you reference your work properly. If you are using any material from the internet or any books from the library, you need to cite the work correctly. Failure to do so will result in possible cases of Academic Misconduct. 4. Please do not share your work with other students. Do not give anyone your files to have a look. SafeAssign will pick up collusion, but keep in mind the percentages for Collusion may not report accurately until all student assignments have been submitted. Both the person copying and the person providing will potentially be held accountable. 5. You can submit a draft assignment through SafeAssign before making the actual submission. 6. If you need any advice or are unsure about referencing, please speak with ATMC Administration for assistance. End of Assignment Assignment Task Part I - PySpark source code (50%) Task I.1: Exploratory data analysis Task I.2: Recommendation engine Task I.3: Classification Task I.4: Clustering Part II –Report (50%)