Perform Residuals Diagnostics and Produce Forecasts Report Writing – Statistics Assignment Help

Assignment Task

Task

The data:

Using the library army of the R statistical software, consider the monthly time senes of the M3 competition with IDs within (1501. 2800), so that the last digit of the series ID matches the last digit of your Student ID For example, if your Student ID is 1456789_ then you should select all the series with IDs finishing at 9, that means- 1509, 1519, 1529, 1539,    2799 Following this procedure, you should end up with a set of 130  monthly time series You should be able to access a single time senes (e.g the time series 1509) using the command a3([1509]1 Note that each time series is split in an in-sample ( /13[[1509] ]$x ) and an out-of-sample ( M [[1509]1$xx ) set of observations Other useful variables include the size of the in-sample ( M3E115091 jSn ), the size of the out-of-sample (equal to the required forecast horizon. t43[ [1509] ]$n ). and the category of the data (micro, macro, industry, demographic, finance or other, /43[[15e9]] ) The length of the out-of-sample set is always 18 months. You are expected to use only the in-sample set in order iproduce statistical forecasts for the out-of-sample set (forecasting horizon equal to 18 penodsimonths) Then, the forecasting performance should be evaluated by comparing the produced forecasts with the withheld out-of-sample set of observations.

Part 2: Batch forecasting

After a short description of your data set, your aim is to produce and evaluate forecasts using:

1. Automatic Exponential Smoothing model selection. One option would be to use the ets() function in the forecast package for R statistical software.

2. Automatic ARIMA model selection One option would be to use the auto. arima() function in the forecast package for R statistical software.

3.  A Model Selection strategy defined by you Using only the in-sample data, propose a suitable strategy in order to select for each series individually the most suitable sets of forecasts from the above two options (Exponential Smoothing and ARIMA) Justify the choice of this model selection strategy over other model selection strategies for forecasting Apply the proposed model selection strategy to the data in order to generate forecasts for the out-of-sample periods. You can consider methodologies such as validation and/or cross-validation that could be applied across a range of methodsimodels. The selection of the validation/cross-validation windows lengths and the number of cross-validation steps should be justified Once the forecasts for the out-of-sample (test data or $xx ) have been produced, you are asked to evaluate these forecasts using at least three appropriate error measures Justify the selection of these error measures over other possible candidates Evaluation should be earned out across time senes and across horizons Compare the accuracy of the proposed combination strategy with that of three suitable benchmark methods (for example. Naive, Damped Exponential Smoothing,etc ) when each of these is applied across all series Critically discuss the performance of each of the five options above In order to achieve marks, you should consider analysing the results for different planning honzons (short-, medium-and long-term), different time series characteristics (trend and/or seasonality) as well as different categories of data (micro, macro, industry, demographic, finance or other).

The structure of the report:

The coursework requires you to document your analysis and critically discuss your chosen design, modelling approaches and your results in a business report suitable for a non-technical audience You are not required to write a general description of the models 0 e. “Exponential Smoothing is “) Instead, the report should document the process of modelling and allow a detailed understanding of your choices You should not
include code within the report; remember, this is a business reportl Your report should have the following sections

Your report should have the tolloteng seclions

• Executive summery
• Introduction Mich should exclude a diw:usiraon on Ducar* sepia cillarscaslingi
• Manual modelling (es desated sets)
• Batch forecasting (as dssOlbed arts)
• Conclusions (including brniebons and menewnel
• References