QBUS6840 Group Assignment
代写预测分析作业 Key information 1.Required submissions: a. ONE written report (word or pdf format, through Canvas- AssignmentsReport submission (group assignment))
Key information
1.Required submissions:
a. ONE written report (word or pdf format, through Canvas- AssignmentsReport submission (group assignment)).
b. ONE code file (Jupyter Notebook “.ipynb” or Python “.py”, through CanvasAssignments- Code submission (group assignment)).
2.The group leader needs to submit both files.
3.Weight: 25% of the total mark of the unit.
4.The full marks of this group assignment are 65 marks. In addition, themaximum bonus marks based on the class forecasting competition are 5 marks.
5.Groups: you should complete this group project in a group of five students.
6.Presentation: please refer to the Presentation Instructions section of this file for more detailed instructions, including the length requirement of the report, font size, etc. To facilitate your report writing process, a Report_Instructions.pdf file is also provided on Canvas.
7.Numbers with decimals should be reported to the four-decimal point.
8.Marking Criteria: please refer to the Marking Criteria section of this file for more detailed instructions.
9.Anonymous marking: as the anonymous marking policy of the University, please only include the student ID of all group members and group ID in the submitted report and code file, and do NOT include your name. The file names of your report and code should follow the following formats respectively, by replacing “123” with your group ID. Example: Group_123_Report, Group_123_Code.
Key rules 代写预测分析作业
- Carefully read requirements of the assignment.
- Please follow any further instructions announced on Canvas.
- You must use Python for the assignment.
- If the training of your model involves generating random numbers, your Python code random seed must be fixed, by using np.random.seed(0).
- Reproducibility is fundamental in data analysis, so that you will be required to submit a code file that generates your results. Not submitting your code will lead to aloss of 50% of the assignment marks.
- Failure to read information and follow instructions may lead to a loss of marks. Furthermore, note that it is your responsibility to be informed of the University of Sydney and Business School rules and guidelines, and follow them.
Background
The unemployment rate is the number of people unemployed divided by the number of people in the civilian labor force. The unemployment rate is reported by the relevant government department in most countries. The unemployment rate can be used as an important indicator by the central bank of the country to determine the health of the economy when setting monetary policy. Further, investors use unemployment statistics to plan future operating activities, see https://www.thebalance.com/unemployment-rate-3305744.
Tasks and Datasets 代写预测分析作业
For this group project, we have obtained the historical unemployment rate data in a country from February 1978 to December 2017. The dataset UnemploymentRate_InSample.csv can be downloaded from the Canvas. The dataset contains information of date (month-year) and unemployment rates.
Your task is to develop a predictive model, trained with UnemploymentRate_InSample.csv, to forecast the monthly unemployment rate from January 2018 to December 2019. Note this is a 24-step-ahead forecast task.
An out-of-sample test dataset which contains the true 2018 and 2019 unemployment rates, named UmemploymentRate_OutofSample.csv in the same format as the in-sample data, is provided on Canvas. They will be used to assess the forecast accuracy of your produced models. Since you should assume the out-of-sample data is completely hidden from your model training/selection process, you must NOT use the out-of-sample test dataset in your model training/selection process. Otherwise, your model training process will be treated as having critical issues and you will receive heave penalty on the methodology and forecasting results, no matter how good your forecasting results are.
In other words, the out-of-sample test dataset should be only used to evaluate your forecast accuracy (details to be shown later). 代写预测分析作业
Please note the assignment tasks are deliberately designed to be open-ended questions. This gives more freedom for you to explore a good solution and is similar to the situations that you might encounter in the real world.
You need to prepare a report for this assignment. The purpose of the report is to describe, explain, and justify your solutions with polished presentation. Be concise and objective. Find ways to say more with less.
You MUST submit your Python code used to produce the results in your report. Suggested Report Outline:
1.(2 marks)
At the beginning (the first line) of your report, you should report your best out-of-sample forecasting result, by stating: “The best out-of-sample forecasting Root Mean Squared Error of our group is: ……”. Please note the markers will run your code and check whether your reported results can be produced/replicated.
Reporting false results deliberately can result in an up to 30% penalty of the assignment marks.
2.(5 marks)
Introduction. Write a few paragraphs stating the business problem andsummarising your works, etc. Use plain English and avoid technical language as much as possible in this section (it should be for the general audience).
3.(10 marks) 代写预测分析作业
Data pre-processing and exploratory data analysis (EDA). Write python program to clean the data, e.g., checking/deleting incomplete information if any, making sure data is complete, or transforming the data if needed, etc. It is up to you on whether/how to transform the data so that the resulting dataset can be well incorporated in training your chosen models.
Conduct initial analysis of the time series by plotting them or do what you can to reveal any patterns. Summarise what you have revealed or observed. In your report, carefully present your EDA procedure and findings.
4.(40 marks)
Methodology and forecasting results. In your report, you should present the details of three different models. The three models should be different types of models. For example, ARIMA(1,1,0) and ARIMA(2,0,1) are the same type of models. ARIMA and Seasonal ARIMA models will be counted as different types of models. Simple Exponential Smoothing, Trend Corrected Exponential Smoothing, and Seasonal Holt-Winters Smoothing models will be counted as different types of models. Additive and Multiplicative Seasonal Holt-Winters Smoothing models will be counted as different types of models. Neural Networks Autoregression and Recurrent Neural Networks models will be counted as different types of models.
The details of the methodology/model should include: 代写预测分析作业
your rationale, how you train your models, model selection process, some interpretations, your findings and justifications of your choices. You can try models that are not covered in our unit. However, for the three models presented, at least two models should be the models that we have covered in the lecture. The types of models could be the Moving Average, Decomposition method, Exponential Smoothing, ARIMA, Neural Networks Autoregression Model, Recurrent Neural Networks, Forecasting Simple Average, Forecasting Combination, etc. This is your choice.
Then you report the out-of-sample forecasting results of your three presented models. In particular, your best model’s out-of-sample forecasting result should be presented at the beginning of the report, as mentioned in above point (1). This best model’s result will decide the forecast competition bonus marks for your group, refer to the Marking Criteria section later for more details.
Calculation of the out-of-sample forecasting results.
You need to use your trained models to predict the 24 unemployment rates of 2018 and 2019. Please note that this is a 24–step-ahead forecast, since we assume you are in December 2017 (time stamp) and have no knowledge about 2018 and 2019. Therefore, as mentioned you should assume the out-of-sample data is completely hidden from your model training/selection process, and you must NOT use the out-of-sample test dataset in your model training/selection process.
5.(3 marks)
Final analysis, conclusion, limitations and future steps (non-technical).
6.Appendix.
In the appendix section, you MUST include three meeting minutes using the provided Minutes Template on Canvas. More detailed instructions are also given below. You can also put any other materials that you see appropriate into the Appendix section. The Appendix will NOT be counted into the length of the main
report and there is no page limit for the Appendix.
Meeting Minutes 代写预测分析作业
- Your group is required to submit three meeting minutes which are to be attached to the report as the Appendix. Your group should use the Minutes Template on Canvas for preparing agendas and meetings minutes.
- You should document at least three meeting minutes for this group assignment, using the template provided. Each minute should at least record the following information:
o Meeting dates/time/duration;
o Key points of the process of discussion, such as who said/did what);
o Action list, responsible member(s), task due time, etc. It is crucial that you clearly document the actions and works for each member during each meeting;
o Review/group judgement on the quality of individually completed/responsible tasks. The purpose of this is to infer whether a member is doing his/her share of jobs.
In case of a problem raised within a group, we will request minutes of all group meetings. We will make an individual adjustment to the group mark, if there is sufficient evidence shows that a student has done significantly less works than other members. If a student has truly done very little works, a mark of 0 will be awarded for the student.
Marking Criteria 代写预测分析作业
The full marks of this group project are 65 marks, including 60 marks for the report and 5 marks for the presentation. In addition, the maximum bonus marks based on the class forecasting competition are 5 marks. The details are shown below:
-
The content in your report Group_123_Report is worth 60 marks
(with suggested report structure and mark break down as above in the Suggest ReportOutline section):
o Focus on the appropriateness of the chosen forecasting method and provide full explanation and interpretation of any results you obtain in your report. Output without explanation will receive 0 marks.
o Describe your data analysis procedure in detail: how the data pre-processing is completed, how the EDA is done, what and why these models are used, how the models are trained, the model selection process, some interpretations, your findings and justifications of your choices. The descriptions should be detailed enough so that other data scientists, who are supposed to have background in your field, understand and are able to implement your works.
o Clearly and appropriately present any relevant graphs and tables.
o You may insert small section of your code into the report for better interpretation when necessary.
-
The Python implementation. 代写预测分析作业
The main program file should be named asGroup123_code.ipynb (or Group123_code.py). Your program must be runnable and your out-of-sample forecasting RMSE results must be replicable. Reporting false results deliberately can result in an up to 30% penalty of the assignment marks.
The idea is that, when the marker runs your Group_123_Code.ipynb (or Group_123_Code.py), with the in-sample train data
UnemploymentRate_InSample.csv and out-of-sample test data
UmemploymentRate_OutofSample.csv in the same folder as the Python file, the marker expects to see the same out-of-sample RMSE value as you reported. The code file should contain sufficient explanations so that the marker knows how to run your code.
-
Presentation is part of the assessment.
The marker will assign 5 marks forpresentation. The detailed instructions are shown in the following PresentationInstructions section.
-
We will allocate a maximum of 5 bonus marks for the forecast competition among the groups. 代写预测分析作业
Groups will receive marks according to the rank of your best out-of sample forecast RMSE value (the value that you reported at the beginning (the first line) of your report; the smaller the better), according to the following rules:
o If the out-of-sample forecast RMSE of your forecast is within top 5 percent in the class, then the full 5 bonus marks (for each student in the group) will be awarded;
o If the out-of-sample forecast RMSE of your forecast is between 5.1 percent (using one decimal rounding) and 20 percent in the class, then 3 bonus marks (for each student in the group) will be awarded;
o If the out-of-sample forecast RMSE of your forecast is between 20.1 percent (using one decimal rounding) and 50 percent in the class, then 1 bonus mark (for each student in the group) will be awarded;
o Otherwise, 0 bonus marks will be awarded.
Presentation Instructions 代写预测分析作业
- Your report should be provided as a word or pdf document.
- Each group should submit one report and one code file.
- To facilitate your report writing process, a Report_Instructions.pdf file is provided.
- The report should be NOT more than 15 pages (excluding Appendix), with font size no smaller than 11pt. The page limit applies to all the content in your report, such as text, figures, tables, small sections of inserted codes, etc, but excluding the Appendix. A violation of this rule will incur penalty on the presentation marks.
- You do NOT need to include a cover page.
- Numbers with decimals should be reported to the four-decimal point.
- You report should:
o Include sections as suggested in Suggested Report Outline section.
o Include all the methodology details and steps as mentioned above.
o Demonstrate an understanding of the relevant principles of predictive analytics approaches used.
o Clearly and appropriately present any relevant figures and tables.
- Your group is required to submit three meetings minutes. Your group should use the Minutes Template provided on Canvas to prepare agendas and meetings minutes. Not providing the meeting minutes will incur penalty on the presentation marks.
- Later, the unit coordinator will collect peer feedback on the performance of each group member. Therefore, it is crucial that each group member is contributing genuinely to the group assignment.
更多代写：cs美国网课代修 代考靠谱吗 英国经济学代上网课 report报告结构代写 PPT演讲稿范例代写 好写的论文题目
