4 85 TRANSFERRING TARGETED MAXIMUM LIKELIHOOD ESTIMATION INTO SPORT SCIENCE dataset is available online 6. Note that in the remainder we work with a min-max normalized, bounded version of Y ∈ [0,1]. While this is not relevant for the initial steps of the roadmap, the boundedness of Y will become important in the later steps (specifically the estimation step). Figure 2. Number of substitutions in the second half per 5-minute period. Figure 3. Difference in the total distance when a substitution took place in the previous period or not ( A). 6 Available at https://github.com/dijkhuist/Entropy-TMLE-Substitutions/tree/main/Data.
RkJQdWJsaXNoZXIy MjY0ODMw