Part 1
Paul Augustinack
IS 340
Final Project
Design View:
View Query:
Decision Tree:
Analysis:
For this prediction, I chose to look at the player seed, player aces, first serve points won, and the total number of service games by played by player. Using the data above in the decision tree we can determine that the order of significance is player seed, the total number of service games, player aces and lastly the first serve points won. This makes sense because the player seeds are developed over the course of the season and players are ranked based on their previous tournaments in the season. Next, the number of saved games is important because the more games played, means the more games you have played in the tournament, or the closer you are to winning the tournament. Player aces are important, however they don't determine a match. It makes sense for this to be less significant than the seeds or games saved. Lastly, the first match won. This characteristic isn't as significant because it is the first of numerous matches they will need to win to win the US Open. Taking a look at the probability of each statistic using the decision tree we can also see why the order of the significance is the order described above. The player seed probability from 1-32 has a 71% chance of winning over anyone that is not ranked. If the person is ranked from 1-4 the probability moves up to 84%. Looking at the service games next, we can see that if you're missing service games your probability increases to 95% compared to if you are probability of winning drops to 78%. With player aces, the significance drops. If the player is not missing player aces his probability is up towards 80% versus 60% if he was missing aces. The first serve points is far less significant that depends a lot on the other attributes we looked at. The probability ranges from 98% all the way down to 28%.
Design View of Query #2:
View of Query #2:


Comments
Post a Comment