Dqs for Week 3 and 4 in Res342

Analyzing the residuals tells us if we have captured all the variability in the data with our model. Ideally, the residuals are randomly distributed, indicating that the model has fully captured the relationship between the dependent and independent variables.

What are the components of a time series?

The time is the independent variable, and the measured quantity is the dependent variable.

What external factors might affect each of the different components?

Many things can change over time, such as environmental conditions or other biasing factors.

What is the purpose of forecasting?

Predicting the future under some assumed conditions.

Why are forecasts not always correct?

Often the models are inexact, and other times the models are used beyond their range of validity.

What are the limitations of forecasting?

Forecasting is limited by the ability of the model to capture all important driving factors, and by noise and uncertainty in the measurements.

Why is the trend commonly thought to be the most important component of forecasting?

Linear trends are the most common because this is the most simple model possible. Also, nearly any relationship can be well modeled as linear over a small enough range.

How is the linear trend method related to linear regression?

The trend is the slope of the line determined by linear regression.

Why are simple moving averages and weighted averages commonly used forms of forecasting?

These are easily computed, and effective in reducing the effects of random noise in the data.

What is an index number as relates to time series analysis?

The index number is simply the sequential number assigned to a particular measurement pair.

In a regression of a time series that states data as a function of calendar year, what requirement of regression is violated?

The independent variable is not measured on equivalent scales – some years are longer than others because of leap years.

How might one...