Basicstepsinforecasting_111 x
Develop a statistically based forecast for the future trend of the essential influencing factor (Covid Pandemic) of a chosen industry or business in Slovenia. The forecast can be monthly or quarterly from 2019 indices forecast on the 2020 indices in the chosen sector’/industry exports or production. You can use the impact of Covid 19 on the sector to create the forecast using graphs and a triple exponential smoothing model. You can Choose
1. Metal Manufacturing Industry Exports.
or
2. Motor Vehicles Manufacturing Industry Exports in Slovenia.
Data Sources to choose from
https://trade.nosis.com/en/Comex/Import-Export/Slovenia/electric-motors–electric-motors-and-generators-excluding-generating-sets/SI/8501
https://pxweb.stat.si/SiStatDb/pxweb/en/20_Ekonomsko/20_Ekonomsko__17_rudarstvo_predel__17011_ind_proiz/1701101S.px/
https://data.worldbank.org/indicator/NV.IND.TOTL.ZS?view=chart
Basic Outline for the forecasting:
1. Problem definition
– Be specific
a)
Expert opinions
2. Exploratory Analysis
a)
Graphical Analysis
(Pictures)
b)
Descriptive statistics
(Mean, Standard deviation, Correlation, Min, Max…)
c)
Decomposition analysis
(trend, seasonality, cyles, outliers-noise)
Just
https://www.wessa.net/rwasp_decomposeloess.wasp#output
put data here it will do it automatically graphs of seasonality and trend and cycles etc. Then you have to explain it
3. Model Choice
a)
Do the (explicit/implicite) assumptions of the chosen model fit my data?
b) (triple exponential smoothing – multiplicative from the script)
4. Fitting the model
a)
Parametrize the model in a way that the historical data is fitted (error measures)
It has to be (triple exponential smoothing – multiplicative from the script) and finding error can be found on excel its you triple smoothing – real data
5. Model Use and Backtesting (Evaluation)
a)
Forecasting errors
b) Using SOVLVER in Excel