forecasting supply chain requirements

Question 1

distinguish between qualitative and quantitative forecasting;

Answer 1

qualitative forecasting techniques.
1. The Delphi Method – a panel of experts come up with a forecast based on wisdom
and insight.
2. Jury of executive option – Subjective opinions of managers and executives are
combined to develop a forecast.
3. Sales force composite – forecasts are developed by sales persons per region and
combined at a district or national level.
4. Consumer market survey – surveys are conducted with customers. An analysis of
the survey results is used to determine future purchasing plans and forecasts are
developed accordingly.
Quantitative forecasting techniques.
Time series models make use of time to predict demand patterns. These models make
use of historical demand data (historical sales information) to generate a forecast.
Explanatory models are also known as regression models and make use of related
variables (causal factors) to predict values.

Question 2

distinguish between time series forecasting and explanatory forecasting;

Answer 2

1. Causal forecasting models the cause-and-effect relationships between variables, while time series forecasting relies on historical data patterns to predict future values. 2. Time series forecasting is limited to observed patterns in the data, whereas causal forecasting takes into account specific factors or variables that influence the forecasted variable

Question 3

identify and describe some qualitative methods of forecasting;

Answer 3

Quantitative forecasting techniques.
Time series models make use of time to predict demand patterns. These models make
use of historical demand data (historical sales information) to generate a forecast.
Explanatory models are also known as regression models and make use of related
variables (causal factors) to predict values.

Question 4

establish appropriateness of time series models;

Question 5

apply relevant time series models to data;

Answer 5

mean absolute deviation. sum of absolute deviations divide buy the number of observations.
2.7.1.2 Mean Square Error (MSE)
The Mean Square Error is calculated by averaging the squared deviations and places more weight on the large errors than on small errors.
2.7.1.3 Mean Absolute Percentage Error (MAPE)
MAPE displays the average absolute error as a percentage.
Average of the absolute percentage errors divide by the Number of observations

Question 6

what are some of the features of forecasting

Answer 6

forecast are always wrong.
it is important to include an estimate of error in the calculation.
forecasts are usually more accurate for a group of items then single items.
the past is an good estimate of the future.

Question 7

what are elements of a good forecast

Answer 7

Forecasts should be timely.
Forecasts should be credible and reliable.
Document and communicate all assumptions made and conditions.
Make use of relevant and applicable forecast techniques for the type of products.
The forecast process must be cost effective.
Forecasts should address uncertainty by including a measure of forecast error.

Question 8

steps in the forecasting process

Answer 8

identify the objective of the forecast.
identify what to forecast.
determine if the forecast is short term , medium term or long term.
gather the historical data to be used in the forecast.
select the best forecast method
generate the forecast.
implement the results
track the results and measure accuracy.

Question 9

techniques for time series data

Answer 9

2.8.1 Naïve Forecasting
The Naïve Forecasting method is very simple. Basically it assumes that the next period forecast is equal to the current period’s actual value (actual sales)
simple moving average: This method simply calculates the average over a period of time (say three periods of actual demand for example) and uses those periods of information to forecast the next period’s demand.Month 1 + Month 2 + Month 3 + Month 4 + Month 5 + Month 6 divide by Number of weeks.
weighted moving average forecast. A weighted moving average is calculated by multiplying each period by a weighting factor, and dividing the resulting product by the sum of all weighting factors.

Question 10

steps to forecast seasonality

Answer 10

Step 1: Calculate the forecast demand for the known observations using linear regression
or Holt’s method.
Step 2: Calculate the actual sales as a percentage of the trend by dividing the actual
sales by the forecast sales for each time period.
Step 3: Calculate the average seasonal index for each season by adding up the values
calculated in step 2 for that season and dividing by the number of years.
Step 4: Multiply the forecast by the average seasonal index for the corresponding
season. This will produce a forecast adjusted for seasons.