TZstatsADS / Spr2017-proj5-grp1

Spr2017-proj5-grp1 created by GitHub Classroom
1 stars 3 forks source link

Baseline on NBA Predictive Model #1

Open amandazhang opened 7 years ago

amandazhang commented 7 years ago

After first meeting today, we decided to using machine learning tools to build a prediction model for 2016-2017 NBA Playoffs.

TO do list:

amandazhang commented 7 years ago

Feature Selection model with take in 2 d dimensional vectors, representing the two teams playing in the tournament. Each d dimensional vector will contain information about the team for that year. Below are the features we end up using. Regular Season Wins Points per game season average Number of 3's per game Turnovers per game average Assists per game average Conference Championship - binary label Rebounds per game average Steals per game average Whether the team is home or away or neutral (labels -1, 0, and 1)

amandazhang commented 7 years ago

image

amandazhang commented 7 years ago

Western Conference: Golden State vs Portland, LAC vs Utah, SA vs MEM, Hou vs OKC

Eastern Conference: BOS vs CHI, WAS vs ATU, CLE vs IND, TOR vs MIL

amandazhang commented 7 years ago

Update on playoff game played till 4/19 image