Reference no: EM132532528
COSC 2670 Practical Data Science with Python - RMIT University
General Requirements
This section contains information about the general requirements that your assignment must meet. Please read all requirements carefully before you start.
• You must do all modelling in IPython or Jupyter Notebook (in Anaconda).
You must include a plain text file called "readme.txt" with your submission. This file should include your name and student ID, and instructions for how to execute your submitted script files. This is important as automation is part of the 6th step of data science process, and will be assessed strictly.
• Parts of this assignment will include a written report, this must be in PDF format.
Please ensure that your submission follows the file naming rules specified in the tasks below. File names are case sensitive, i.e. if it is specified that the file name is gryphon, then that is exactly the file name you should submit; Gryphon, GRYPHON, griffin, and anything else but gryphon will be rejected.
Part 1: Retrieving and Preparing the Data
This assignment will focus on data modelling, and you can choose to focus on one ap- proach: Classification or Clustering.
For this assignment, you need to select one dataset from the following options, and then work on it:
1. Activity Recognition from Single Chest-Mounted Accelerometer Data Set. More details can be found from the following UCI webpage about this dataset
2. BLE RSSI Dataset for Indoor localization and Navigation Data Set. More details can be found from the following UCI webpage about this dataset
3. Mice Protein Expression Data Set. More details can be found from the following UCI webpage about this dataset
Being a careful data scientist, you know that it is vital to set the goal of the project, then thoroughly pre-process any available data (each attribute) before starting to analyse and model it. In your report in Part 4, You need to clearly state the goal of your project, and the design/steps of pre-processing your data. Please ensure you understand the data you selected, including the meaning of each attribute.
Part 2: Data Exploration
Explore the selected data, carrying out the following tasks:
Explore each column (or at least 10 columns if there are more than 10 columns), using appropriate descriptive statistics and graphs (if appropriate). For each ex- plored column, please think carefully and report in your report in Part 4): 1) the way you used to explore a column (e.g. the graph); 2) what you can observe from the way you used to explore it.
(Please format each graph carefully, and use it in your final report. You need to include appropriate labels on the x-axis and y-axis, a title, and a legend. The fonts should be sized for good readability. Components of the graphs should be coloured appropriately, if applicable.)
Explore the relationship between all pairs of attributes (or at least 10 pairs of at- tributes, if there are more in the data), and show the relationship in an appropriate graphs. You may choose which pairs of columns to focus on, but you need to gen- erate a visualisation graph for each pair of attributes. Each of the attribute pair should address a plausible hypothesis for the data concerned. In your report, for each plot (pair of attributes), state the hypothesis that you are investigating. Then, briefly discuss any interesting relationships (or lack of relationships) that you can observe from your visualisation.
Please note you do not need to put all the graphs in your report, and you only need to include the representative ones and/or those showing significant information.
Part 3: Data Modelling
Model the data by treating it as either a Classification or Clustering Task, depending on your choice.
You must use two different models (i.e. two Classification models, or two Clustering models), and when building each model, it must include the following steps:
• Select the appropriate features
• Select the appropriate model (e.g. DecisionTree for classification) from sklearn.
• If you choose to do a Classification Task,
- Train and evaluate the model appropriately.
- Train the model by selecting the appropriate values for each parameter in the model. You need to show how you choose this values, and justify why you choose it.
• If you choose to do a Clustering Task,
- Train the model by selecting appropriate values for each parameter in the model.
Show how do you choose this value, and justify why you choose it (for example, k in the k-means model).
- Determine the optimal number of clusters, and justify
- Evaluate the performance of the clustering model by:
∗ Checking the clustering results against the true observation labels Constructing a "confusion matrix" to analyse the meaning of each cluster by looking at the majority of observations in the cluster. (You can do this
by using a pen and a piece of paper, as we did in Practical Exercise; if you prefer, you can also explore how to do this step directly in IPython.)
After you have built two Classification models, or two Clustering models, on your data, the next step is to compare the models. You need to include the results of this comparison, including a recommendation of which model should be used, in your report (see next section).
Part 4: Report
Write your report and save it in a file called report.pdf, and it must be in PDF format, and must be at most 12 (in single column format) pages (including figures and references) with a font size between 10 and 12 points Penalties will apply if the report does not satisfy the requirement. Remember to clearly cite any sources (including books, research papers, course notes, etc.) that you referred to while designing aspects of your programs.
Your report must have the following structure:
• A cover page, including
- Statement of the solution representing your own work as required
- Title
- Author Information
- Affiliations
- Contact details
- Date of report
• Table of Content
• An abstract/executive summary
• Introduction
• Methodology
• Results
• Discussion
• Conclusion
• References
Please revisit the relevant slides in Week1 lecture if needed.
Part 5: Presentation
• The presentation should
- explain the goal of the project.
- briefly describe your chosen data set.
- describe the data preparation steps.
- state the hypotheses/questions that you were investigating.
- explain what the modelling steps are, and what the results are.
- show the final conclusion and recommendation.
• The presentation should be no more than 5 minutes.
• Your presentation slides should be:
- Microsoft PowerPoint slides (with audio inserted for each slide by using: Insert
- > Audio - > Record Audio).
- or you can create your own presentation slides (e.g. PDF version) and please submit your own recording of your presentation as well.
Attachment:- Practical Data Science with Python.rar