Critical Thinking

Critical Thinking

1 .
 
a.
What characteristics define a time series?
b.
Which of the following are examples of time series data?
  1. The vital statistics of a cancer patient taken right before a major surgery
  2. The monthly expenses of a small company recorded over a period of five years
  3. Daily temperature, rainfall, humidity, and wind speeds measured at a particular location over a few months
  4. Student final grades in all sections of a course at a university
2 .
Consider the graph in the figure as shown below.
a.
Is the trend curve generally increasing, decreasing, or staying level over the entire time period?
b.
The data set USATemps1961-2023.csv contains the time series from which the graph was created. Find the moving average trendline with period 10 for the data.

Line graph titled “Temperature deviation from baseline in the United States (in degrees Celsius).” The graph shows fluctuations of a value over time. Y-axis ranges from -1 to 2.5, x-axis from 1961 to 2021. The line exhibits a generally upward trend with significant peaks and troughs. Figure 5.20
3 .
 
a.
What are the characteristics of white noise? Why is it important that the residuals of a time series model be white noise?
b.
Determine which of the following graphs most likely represents white noise.
Four white noise line graphs labeled I, II, III, and IV. Y axis labeled Value and X axis labeled Time. Y axis ranges from -6 to 6. X axis ranges from 0 to 200. Graph I displays random fluctuations around 2. Graph II displays random fluctuations that start around 0 but have significant variance around 100 seconds. Graph III displays random fluctuations around 0. Graph IV displays a more wavelike variance of random fluctuations.
4 .
Four time series models were produced to model a given dataset. Various measures of error were run on each of the models, the results of which are shown in the table.
MAE RMSE MAPE
Model 1 103.1 131.2 0.097
Model 2 251.7 207.9 0.136
Model 3 110.7 129.0 0.115
Model 4 89.8 125.3 0.191
Table 5.8 Measures of Error
a.
Identify the most accurate model based only on:
  1. MAE
  2. RMSE
  3. MAPE
b.
Which model is likely to produce the least accurate forecasts overall, and why?
c.
Which model is likely to produce the most accurate forecasts overall, and why?