In the same theme,MLB.
I will show how to use Python to train a randomforest and SVM classifier.
Our target is predicting whether the team will enter the next year playoff or not. Since the playoff qualification depends on the same year’s season game win rate. If we predict the same year playoff is pointless. So I make the variable “palyoff” to “next year playoff”. I will use the module “sklearn” to finish my work.

Import module

import numpy as np
import pandas as pd
from sklearn import cross_validation, ensemble, preprocessing, metrics 

Read predict data

data = pd.read_csv('C:/Users/User/OneDrive - student.nsysu.edu.tw/Educations/2019暑期實習/工研院/資料科學_mlb/mlb_predict.csv')

Build Train and Test data set

I choose the last article regression model variable. That will become more sensible since the playoff relate with win rate.
And I cut the data on 7:3 to train set and test set.

mlb_X = pd.DataFrame([data["Bat"],data["AB"],data["HR_x"],data["SO_x"],data["OPS+"],data["GDP"],
                      data["IBB_x"],data["#P"],data["PAge"],data["tSho"],data["SV"],data["IP"],
                      data["HBP_y"],data["ERA+"],data["WP"]]).T
mlb_y = data["Playoff_next"]
train_X, test_X, train_y, test_y = cross_validation.train_test_split(mlb_X, mlb_y, test_size = 0.3)

Random Forest

Build the random forest model

I chose to use 100 trees to build the model this time.

forest = ensemble.RandomForestClassifier(n_estimators = 100)
forest_fit = forest.fit(train_X, train_y)

Predict on random forest model

test_y_predicted = forest.predict(test_X)
accuracy_rf = metrics.accuracy_score(test_y, test_y_predicted)
print(accuracy_rf)

0.7350427350427351

It shows that the accuracy is 0.7350427350427351.

SVM

Build SVC model

I didn’t set any parameter this time. All is default.

from sklearn import cross_validation, svm, preprocessing, metrics
svc = svm.SVC()
svc_fit = svc.fit(train_X, train_y)

Predict on random forest model

# Predict
test_y_predicted = svc.predict(test_X)
accuracy_svm = metrics.accuracy_score(test_y, test_y_predicted)
print(accuracy_svm)

0.7521367521367521

It shows that the accuracy is 0.7521367521367521.