notes of course “Introduction to Data Science” from RWTH-Aachen in semester winter 19/20, Professor van der Aalst, Willibrordus

Two questions

Performance Problems & Compliance Problems

• Reporting : What happened?
• Diagnosis: Why did it happen?
• Prediction: What will happen?
• Recommendation: What is the best that can happen?

Conformance checking

Fitness of the model

$\text { fitness }(\sigma, N)=\frac{1}{2}\left(1-\frac{m}{c}\right)+\frac{1}{2}\left(1-\frac{r}{p}\right)$ fitness (σ,N)=21​(1−cm​)+21​(1−pr​)

Counting tokens while replaying:

• p = produced tokens
• c = consumed tokens
• m = missing tokens
• r = remaining tokens

At any time: $p+m\geq c \geq m$p+m≥c≥m (You are not adding missing token that you do not need)
At the end: $r=p+m-c$r=p+m−c
Not necessarily p = c
For instance:

Initialization: p = 1
Finalization: c = c + 1

Fitness at the log level
$\begin{aligned} \text { fitness }(L, N)=& \frac{1}{2}\left(1-\frac{\sum_{\sigma \in L} L(\sigma) \times m_{N, \sigma}}{\sum_{\sigma \in L} L(\sigma) \times c_{N, \sigma}}\right)+frac{1}{2}\left(1-\frac{\sum_{\sigma \in L} L(\sigma) \times r_{N, \sigma}}{\sum_{\sigma \in L} L(\sigma) \times p_{N, \sigma}}\right) \end{aligned}$ fitness (L,N)=​21​(1−∑σ∈L​L(σ)×cN,σ​∑σ∈L​L(σ)×mN,σ​​)+frac12(1−∑σ∈L​L(σ)×pN,σ​∑σ∈L​L(σ)×rN,σ​​)​

Generating supervised learning problems

Tooling

• ProM
• Disco
• celonis
• Lexmark
• minit
• myinvenio