Penerapan Data Mining Prediksi Hasil Taruhan Bola

Data

Algoritma

Generalize Linear

Hasil

Model Metrics Type: BinomialGLM
Description: N/A
model id: rm-h2o-model-generalized_linear_model-850838
frame id: rm-h2o-frame-generalized_linear_model-538670
MSE: 0.24158409
R^2: 0.03358188
AUC: 0.60356915
logloss: 0.67600983
CM: Confusion Matrix (vertical: actual; across: predicted):
YA Tidak Error Rate
YA 35 342 0.9072 = 342 / 377
Tidak 11 373 0.0286 = 11 / 384
Totals 46 715 0.4639 = 353 / 761
Gains/Lift Table (Avg response rate: 50.46 %):
Group Cumulative Data Fraction Lower Threshold Lift Cumulative Lift Response Rate Cumulative Response Rate Capture Rate Cumulative Capture Rate Gain Cumulative Gain
1 0.01051248 0.661410 1.238607 1.238607 0.625000 0.625000 0.013021 0.013021 23.860677 23.860677
2 0.02102497 0.655460 1.486328 1.362467 0.750000 0.687500 0.015625 0.028646 48.632813 36.246745
3 0.03022339 0.651493 1.132440 1.292459 0.571429 0.652174 0.010417 0.039063 13.244048 29.245924
4 0.04073587 0.638516 1.238607 1.278562 0.625000 0.645161 0.013021 0.052083 23.860677 27.856183
5 0.05124836 0.632495 1.238607 1.270366 0.625000 0.641026 0.013021 0.065104 23.860677 27.036592
6 0.10118265 0.612644 1.147341 1.209652 0.578947 0.610390 0.057292 0.122396 14.734101 20.965233
7 0.15111695 0.599116 1.355948 1.257994 0.684211 0.634783 0.067708 0.190104 35.594846 25.799366
8 0.20105125 0.586477 0.990885 1.191653 0.500000 0.601307 0.049479 0.239583 -0.911458 19.165305
9 0.30091984 0.561124 1.329873 1.237525 0.671053 0.624454 0.132813 0.372396 32.987253 23.752502
10 0.40078844 0.538306 1.069113 1.195560 0.539474 0.603279 0.106771 0.479167 6.911321 19.556011
11 0.50065703 0.517563 1.016961 1.159934 0.513158 0.585302 0.101563 0.580729 1.696135 15.993411
12 0.60052562 0.491642 1.043037 1.140494 0.526316 0.575492 0.104167 0.684896 4.303728 14.049394
13 0.70039422 0.460069 0.599746 1.063389 0.302632 0.536585 0.059896 0.744792 -40.025356 6.338923
14 0.80026281 0.426700 1.095189 1.067358 0.552632 0.538588 0.109375 0.854167 9.518914 6.735769
15 0.90013141 0.387787 0.704050 1.027049 0.355263 0.518248 0.070313 0.924479 -29.594984 2.704912
16 1.00000000 0.142783 0.756202 1.000000 0.381579 0.504599 0.075521 1.000000 -24.379797 0.000000
null DOF: 760.0
residual DOF: 755.0
null deviance: 1054.9056
residual deviance: 1028.887
GLM Model (summary):
Family Link Regularization Number of Predictors Total Number of Active Predictors Number of Iterations Training Frame
binomial logit Elastic Net (alpha = 0.5, lambda = 1.708E-4 ) 5 5 2 rm-h2o-frame-generalized_linear_model-538670
Scoring History:
timestamp duration iteration negative_log_likelihood objective
2019-01-12 12:18:58 0.000 sec 0 514.47600 0.67612
2019-01-12 12:18:58 0.002 sec 1 514.44348 0.67609
H2O version: 3.8.2.6-rm7.6.1

Evaluasi Algortima

PerformanceVector:
accuracy: 58.07% +/- 4.63% (mikro: 58.08%)
ConfusionMatrix:
True: YA Tidak
YA: 196 138
Tidak: 181 246
precision: 57.77% +/- 4.54% (mikro: 57.61%) (positive class: Tidak)
ConfusionMatrix:
True: YA Tidak
YA: 196 138
Tidak: 181 246
recall: 64.10% +/- 6.43% (mikro: 64.06%) (positive class: Tidak)
ConfusionMatrix:
True: YA Tidak
YA: 196 138
Tidak: 181 246
AUC (optimistic): 0.598 +/- 0.070 (mikro: 0.598) (positive class: Tidak)
AUC: 0.598 +/- 0.070 (mikro: 0.598) (positive class: Tidak)
AUC (pessimistic): 0.598 +/- 0.070 (mikro: 0.598) (positive class: Tidak)

Tingkat akurasi hanya 58,07 % Sehingga hasil prediksi bisa juga salah.