Within the context of the examples from your experience/workplace (Information Technology Industry) that you have already described,consider whether using decomposition to forecast would be appropriate. What challenges might you face in applying decomposition? Would you be able to use decompostion for longer term forecasting (consider the dominant components in your series)? Would the seasonally adjusted data be useful in your situation?
If decomposition methods would not be applicable in the forecasting scenario that you have already described, describe one in which this approach would be applicable. Would you use the additive or multiplicative form of the model? Explain.
Provide me a detailed discussion of 1.5 pages.
TIME SERIES DECOMPOSITION .Easy to understand (at least conceptually). .Most individuals dealing with time series assume the presence of trend, influence of a business cycle, seasonality when dealing with monthly or quarterly data, and ever present randomness (or noise). .Among the oldest approaches to time series analysis. TIME SERIES DECOMPOSITION .Assumes data = pattern + error = f (trend-cycle, seasonality, error). .Identifies and separates the sub-patterns (components). .Assumes the series is some function (additive or multiplicative) of the individual components. FOUR COMPONENTS OF TIME SERIES TREND. Long-term growth (or decline). Forces affecting trend include population change, inflation, technological change and productivity increases. CYCLICAL. Irregular wavelike fluctuations of more than one year’s duration. Due to changing economic conditions. SEASONAL. Pattern of change that is completed within one year and repeats itself year after year. IRREGULAR. Fluctuations that may be caused by unpredictable or non-periodic events such as unusual weather, rumors of war, etc. TREND In trend analysis the independent variable is time. Least squares method (simple regression) is most often used to fit trend lines. Exponential trend can be fitted when the series starts slowly and then appears to be increasing at an increasing rate. Growth curves (such as Gompertz or Pearl-Reed logistic) can be fitted when the series exhibits a growth rate that slows down (e.g. market maturity which is common for many industries and product lines). TREND-CYCLE Trend is often mixed up with the cycle. Cycle is the most difficult component to forecast because of the difficulty in anticipating changes in direction (upturns or downturns). Many decomposition methods (e.g., Census II) do not try to separate trend and cycle (identify a trend-cycle component). CYCLICAL BUSINESS INDICATORS Leading Indicators . Provide...
Already registered? Login
Not Account? Sign up
Enter your email address to reset your password
Back to Login? Click here