Data Mining using IBM SPSS Modeler

Association Exercises

After analyzing Bike data,

$1.$ What are two items associated each other?

$2.$ What is support of the first item $($Antecedent$)?$ How can you interpret the results?

$3.$ What percent is the confidence of the first item $($Antecedent$)?$ How can you interpret the results?

$4.$ What percent is the lift of the first item $($Antecedent$)?$ How can you interpret the results?

$5.$ Capture the result of the analysis.

Classification Exercises

After analyzing Bike.Customer.Classification.xlsx from D$2$L

$-$ Choose worksheet: By name: Source Data

$-$ BikeBuyer: change to Target

$-$ Partition: $70\%$ training & $30\%$ testing

$-$ Choose Model: C$5\mathrm{.}0$ in Modeling

$-$ Select $$Use Partitioned data$$ & unselect $$Build model for each split$$

$-$ Choose $$Decision tree$$

$-$ In Analysis, select $$Coincidence matrices$,$Performance evaluation$,$ and $$Evaluation metric$$

$1.$ What are the most important predictors $($top $2)$ to determine purchasing bike?

$2.$ Provide one profile of bike buyers.

$3.$ How much is this analysis accurately classified bike buyers?

$4.$ Is $$Megan Sanchez$$ likely to purchase the bike? How likely is she to purchase or not to purchase the bike?

$5.$ What can the manager do with these results?

Cluster Exercises

After analyzing Store data,

$1.$ How many clusters are most appropriate?

$2.$ What predictor is the most important to make clusters?

$3.$ Compare the characteristics of each cluster.

Cluster $1$ Cluster $2$ Cluster N

Sales Turnover

Store Size

Staff Size

Profit Margin

$4.$ Name each cluster based on its own characteristics.

$5.$ Develop strategies to revamp store performance based on characteristics of each cluster.

$6.$ Capture the result of the analysis.