| Unit title & code |
CIS111-6: Intelligent Systems and Data Mining |
| Assignment number and title | Assignment Task 1: Data Mining Solutions for Direct Marketing Campaign |
| Assessment type | WR |
| Weighting of assessment | 50% |
| Unit learning outcomes |
|
| What am I required to do in this assignment? |
| Task
Students will develop a DM solution for saving the cost of a direct marketing campaign by reducing false positive (wasted call) and false negative (missed customer) decisions. Working on this assignment, students can consider the following scenario. A Bank has decided to save the cost of a direct marketing campaign based on phone calls offering a product to a client. A cost efficient solution is expected to support the campaign with predictions for a given client profile whether the client buys the product or not. Examples of cost-efficient DM solutions for direct marketing are provided on the UCI Machine Learning repository describing a Bank Marketing problem. How students will work Each student is expected to run individual experiments to find an efficient solution and describe experimental results in an individual report. Students could work on the assignment task as: (i) a group manager, (ii) a group member, or (iii) an individual. If students will work in a group, the group manager arranges the comparison and ranking of designed solutions. Method and Technology To design a solution, students will use Data Mining techniques such as Decision Trees. Students are recommended to use R scripting: (i) a Cloud CoCalc, (ii) a development suite RStudio or an RStudio Cloud free for students. Other scripting languages such as Python supported e.g. by Google Colab online platform could be also used. Project Code and Data The assignment project code is available as an R Script. The Bank Marketing data set is available as a csv file. Other data sets (Kaggle or UCI) could also be used. Each solution will be evaluated in terms of the costs of false decisions made on the validation data. Reports will be submitted via BREO. Reports can be prepared with a template. BREO similarity in reports must be < 20% (scripting is not counted). |
| Is there a size limit? |
| 2500 words (task 1) & 2500 words (task 2) |
| What do I need to do to pass? (Threshold Expectations from UIF) |
| 1. Follow a CoCalc tutorial to create an individual account (or install RStudio) (10%)
2. Create an R project containing the given project script and data set (10%) 3. Apply a Decision Tree technique to solve the Bank Marketing task (5%) 4. Work on scripting problems is evaluated and students are expected to demonstrate the knowledge on how to find a solution by using related manuals and google search (10%) 5. Analyse problems of designing a solution which will provide a high prediction accuracy (7%) 6. In total 42% to pass |
| How do I produce high quality work that merits a good grade? |
| 7. Identify a set of parameters required to be adjusted within DM techniques in order to optimise a solution in terms of prediction accuracy (10%)
8. Explain how the parameters of a DM technique influence the prediction accuracy (10%) 9. Run experiments in order to verify the solution designed on the given data set (10%) 10. Analyse and compare the results of the experiments in a group and with results known from the literature (13%) 11. In total 85% |
| How does assignment relate to what we are doing in scheduled sessions? |
| Data Mining techniques and use cases developed in R will be considered during lectures and tutorials. |
| How will my assignment be marked? |
| Your assignment be marked according to the threshold expectations and the criteria on the following page.
You can use them to evaluate your own work and estimate your grade before you submit. |
| # | Weight, % | Lower 2nd – 50-59% | Upper 2nd – 60-69% | 1st Class – 70%+ |
| 1 | Analysis
(30) |
Fair analysis of the basic approaches
|
Relatively good analysis of the relevant literature, mainly covering the state-of-art | Excellent analysis of the relevant literature, fully covering the state-of-art |
| 2 | Design
(40) |
Fair design of a basic solution providing a reasonable performance within a single set of parameters | Design of a solution providing a fair performance in a series of experiments with different sets of parameters
|
Design of a solution providing a performance, competitive to known from the literature, in a series of experiments with different sets of parameters |
| 3 | Conclusion (30) | Fair conclusion on the experimental results obtained within a single set of parameters
|
Conclusion on and comparison of the experimental results obtained within two different sets of parameters | Conclusion on and comparison of the experimental results obtained within multiple sets of parameters, demonstrating a solution which provides a competitive performance |
