PREDICTING THE NCAA BASKETBALL TOURNAMENT WITH
PREDICTINGTHE NCAABASKETBALLTOURNAMENTWITHMACHINELEARNINGA N DL E V A NAJ O N AL OThe Ringer/Getty ImagesRDNTBE WO S K IDH A NO
THETOURNAMENT
MARCH MADNESS 68 teams (4 play-in games) 63 single-elimination games 4 regions, 16 teams each Each team seeded 1-162017 Champion: North Carolina(#1 seed in South region)The Ringer/Getty Images
THE BRACKETTHEBRACKET
19 million brackets submitted to ESPN 263, or 9.2 quintillion, possible brackets Nobody has ever picked a perfectbracket 3 billion in bracket pools annually Warren Buffet offers 1 million for aperfect bracketB R A C K E TO L O G Y/ˌbrakәˈtälәjē/nounthe activity of predicting theparticipants in and outcomes ofthe games in a sports tournament,especially the NCAA collegebasketball tournament.
HISTORICAL RESULTSMatchupHigher Seed Win %1 vs 16100.0%2 vs 1593.9%3 vs 1484.1%4 vs 1380.3%5 vs 1264.4%6 vs 1162.9%7 vs 1061.4%8 vs 950.8%
USING ML TO BUILD BRACKETSFiveThirtyEight’sPrediction EngineKaggle’s March MachineLearning Mania 2017
JonathanAndrew PICKING AB R AC K E T I SR E A L LYHARD!
MODELS
PREVIOUS WORK Mostly statistical modeling Poor choice of features– Seeding– Vegas point spreadH OW W EFIX THIS: Pick only relevant features Analyze a wide range of MLtechniques Build our own model usingrandom forests
THE DATA The Ringer/Getty ImagesPointsField Goals MadeField Goals Attempted3-Pointers Made3-Pointers AttemptedFree Throws MadeFree Throws AttemptedOffensive ReboundsDefensive ReboundsAssistsTurnoversStealsBlocksPersonal FoulsGame Location(Home or Away/Neutral)
CONSTRUCTINGTRAINING DATA 14 chosen statistics computed based onrolling averages of previous games– For teams and their opponents Only include data from 15 most recentgames Game Location represented as binary flag No feature for wins or seedThe Ringer/Getty Images
Adaptive Boosting K-Nearest NeighborsLEARNINGTECHNIQUES Naïve Bayes Neural Network Logistic Regression Support Vector Machine Random ForestsTrained for classification (picking awinner and loser for each matchup)ANDdetermining the likelihood of eachoutcome
RANDOM FORESTS The combination of learning models increases classification accuracy Bagging: average noisy and unbiased models to create a model with low varianceThe random forest algorithm acts as a largecollection of decorrelated decision trees
DECISION TREEFeature 1 of the 1st sample Feature 2 of the Nth sample
RANDOM SUBSETS
RANDOM SUBSETS Decision Tree 2 Decision Tree 1 Decision Tree M
VOTING 1 0 1 1Label as majority vote
METHOD EVALUATION Accuracy Bracket score Log Loss (for probabilities)111where n is the number of games played,yi is the predicted probability of team 1 beating team 2,and yi is 1 if team 1 wins, 0 if team 2 wins.The Ringer/Getty Images
RESULTS
oost.667SVM.683Bayes.698Random Forests.698Regression.762Neural Net.794The Ringer/Getty Images
Bayes610Neural Net650Regression670Random Forests900The Ringer/Getty Images
LOG .579Random Forests.578Neural Net.545Regression.529The Ringer/Getty Images
AlgorithmLog LossAccuracyBracket 570Bayes.579.698610Random Forests.578.698900Neural Net.545.794650Regression.529.762670
REFERENCESImage Credits: Getty Images, The RingerB.J. Coleman, J.M. DuMond, and A.K. Lynch, “Evidence of Bias in NCAA Tournament Selection and Seeding”,Managerial and Decision Economics, vol. 31, 2010.Kaggle.com, March Machine Learning Mania 2017”, mania-2017”, 2017.L.H. Yuan et al, “A mixture-of-modelers approach to forecasting NCAA tournament outcomes”, Journalof Quantitative Analysis of Sports, vol. 11, 2014.M.J. Lopez and G.J. Matthews, “Building an NCAA men’s basketball predictive model and quantifying itssuccess”, Journal of Quantitative Analysis of Sports, vol. 11, 2014.Z. Shi, S. Moorthy, and A. Zimmermann, “Predicting NCAAB match outcomes using ML techniques - someresults and lessons learned”, ArXiv e-prints, “https://arxiv.org/abs/1310.3607”, 2013.
bracket 3 billion in bracket pools annually Warren Buffet offers 1 million for a perfect bracket /ˌbrakәˈtälәjē/ noun the activity of predicting the participants in and outcomes of the games in a sports tournament, especially the NCAA college basketball tournament.
May 02, 2018 · D. Program Evaluation ͟The organization has provided a description of the framework for how each program will be evaluated. The framework should include all the elements below: ͟The evaluation methods are cost-effective for the organization ͟Quantitative and qualitative data is being collected (at Basics tier, data collection must have begun)
Silat is a combative art of self-defense and survival rooted from Matay archipelago. It was traced at thé early of Langkasuka Kingdom (2nd century CE) till thé reign of Melaka (Malaysia) Sultanate era (13th century). Silat has now evolved to become part of social culture and tradition with thé appearance of a fine physical and spiritual .
On an exceptional basis, Member States may request UNESCO to provide thé candidates with access to thé platform so they can complète thé form by themselves. Thèse requests must be addressed to esd rize unesco. or by 15 A ril 2021 UNESCO will provide thé nomineewith accessto thé platform via their émail address.
̶The leading indicator of employee engagement is based on the quality of the relationship between employee and supervisor Empower your managers! ̶Help them understand the impact on the organization ̶Share important changes, plan options, tasks, and deadlines ̶Provide key messages and talking points ̶Prepare them to answer employee questions
Dr. Sunita Bharatwal** Dr. Pawan Garga*** Abstract Customer satisfaction is derived from thè functionalities and values, a product or Service can provide. The current study aims to segregate thè dimensions of ordine Service quality and gather insights on its impact on web shopping. The trends of purchases have
OS 149 11 Basketball Team Photo 1941-1942 OS 149 12 Basketball Team Photo 1942-1943 OS 149 13 Basketball Team Photo 1943-1944 OS 149 14 Basketball Team Photo 1945-1946 OS 150 1 Basketball Team Photo 1946-1947 OS 150 2 Basketball Team Photo 1947-1948 OS 150 3 Basketball Team Photo 1949-1950 OS 150 4 Basketball Team Photo 1952-1953
Nov 03, 2014 · Levida Maxwell 317/917-6356 N/A lmaxwell@ncaa.org Rick Nixon 317/917-6539 317/440-3059 rnixon@ncaa.org Phil Pierce 317/917-6976 317/263-4475 ppierce@ncaa org Sahar Rashid 317/917-6547 317/459-3272 srashid@ncaa org Jenn Rodgers 317/917-6109 317/361-0184 jrodgers@ncaa org Sean Straziscar 317/91
Industry members of the Automotive Council expect to take on more than 7,600 apprentices and 1,700 graduates over the next five years Young people Attracting more young people, especially women, into automotive careers by initiatives such as: – Lending Automotive Council endorsement to several high-quality and effective school, graduate and post-graduate schemes – Automotive Council .
