WRITTEN EXAM UITWERKINGEN MANAGEMENT SCIENCE
Confusion Matrix:
- Sensitivity: hit rate, power, recall, true positive rate
- Specificity: true negative rate
- Accuracy: classification rate, proportion (percentage) correct, simple matching coefficient
- Precision: absolute support, certainty factor, confidence, correct alarm ratio, frequency of
hits, positive predictive value, success ratio
Nearest neighbours:
- To the predictor record are those that are closest on the scatter plot.
- The distance is measured by the Euclidean distance (the straight-line distance) between the
two points:
Measure homogeneity when splitting region:
Evaluates homogeneity of individual regions: Gini Index is a fraction of the
historical records in the region that have outcome i.
- Its values range from 0 to 0.5, where a value of 0 indicates that the region is completely
homogeneous, whereas a value of 0.5 indicates that the region is completely lacking in
homogeneity.
a. This index reaches the minimum value of 0 when all of the historical records in the
region have the same outcome
b. It reaches its maximum value of 0.5 when the region is evenly split between the two
classes
Evaluates overall lack of homogeneity of a group of regions: overall Gini Index
- It is calculated as the weighted average of the Gini Index for each of the regions in the group,
weighted by how many historical records fall into each region.
- Its values range from 0 (all of the regions are completely homogeneous) to 0.5 (all of the
regions are completely lacking in homogeneity).
a. With a split at $89k, the Gini Index for the region on the left is 0 (with 4 records) and
the Gini Index for the region on the right is 0.44 (with 12 records). The Overall Gini
Index
, Regression tree:
- Measures lack of homogeneity of a region by Regression Index: Sum of the square of the
deviations of the output variable values form their mean.
- Measures lack of homogeneity of a group of regions: Overall Regression Index.
Forecasting methods:
1. Last-value forecasting method (naïve method):
a. Ignores all data points in a time series except the last one.
b. Forecast=Last value.
c. When conditions are changing rapidly, it may be that the last value is the only
relevant data point.
2. Averaging forecasting method:
a. Uses all the data points in the time series and simply averages these points.
b. Forecast = Average of all data to date.
c. When conditions are very stable, where even its first few values are considered
relevant for forecasting the next value.
d. It is very slow to respond to changing conditions. It places the same weight on all the
data, even though the older values may be less representative of current conditions
then the last value observed.
3. Moving-average forecasting method:
a. Averages the data for only the most recent time periods. N= Number of recent
periods to consider as relevant for forecasting.
b. Forecast = Average of last n values.
c. When conditions don’t change much over the number of time periods included in the
average. where the last few values are considered relevant for forecasting the next
value.
d. It is slow to respond to changing conditions. It places the same weight on each of the
last n values even though the older values may be less representative of current
conditions than the last value observed.
4. Exponential smoothing forecasting method:
a. It gives the greatest weight to the last value in the time series and then progressively
smaller weights to the older values.
b. , is the smoothing constant
between 0 and 1.
c. Places a weight of a on the last value, on the next-to-last value, on
the next prior value, etc.
i. When =0.5, the method places a weight of 0.5 on the last value, 0.25 on
the next-to-last, 0.125 on the next prior, etc.
d. A small value (say, = 0.1) is appropriate if conditions are relatively stable.
e. A larger value (say, =0.5) is appropriate if significant changes occur frequently.
f. For a time series in the range from somewhat unstable to rather stable, where the
value of the smoothing constant needs to be adjusted to fit the anticipated degree of
stability.
5. Exponential smoothing with trend:
a. Uses the recent values (adjusts exponential smoothing) in the time series to estimate
any current upward or downward trend in these values.
b. Trend= Average change from one time-series value to the next.