#Naive Bayes and Neural Networks
So far I tried Naive Bayes and Neural Networks. They both fail for the same reason -- the # of bikes taken out can be anywhere from 0 to 997. It's too broad of a classification to make.
=======
- RMSLE sigmoid = 1.53555698918
- RMSLE linear = 2.85748622823
- RMSLE linear with C=5 = 2.55072279164
- RMSLE linear with C=5 and loss=l1 and penalty=l2 = 1.9382315027
- RMSLE linear with C=10 = 1.50648948479
- RMSLE linear with C=100 = 2.31095759634
- RMSLE linear with C=1000 = 1.98174353979
- RMSLE rbf = 1.53038228682
- RMSLE rbf with C=5 = 1.50591263209
=======
- RMSLE sigmoid = 1.51505901557
- RMSLE linear = 2.66486578715
- RMSLE poly =
- RMSLE rbf = 1.43826321753
- RMSLE rbf with C=5 =
======= #Nearest Neighbors (kd_tree)
- RMSLE nearest neighbors with neighbors=5 = 1.30106865675
- RMSLE nearest neighbors with neighbors=7 = 1.30981933469
- RMSLE nearest neighbors with neighbors=8 = 1.31414497154
- RMSLE nearest neighbors with neighbors=10 = 1.3192726732
- RMSLE nearest neighbors with neighbors=5, leaf=70 = 1.29803419224
- RMSLE nearest neighbors with neighbors=5, leaf=70, p=1 = 1.27996796419
- RMSLE nearest neighbors with neighbors=6, leaf=100, p=1 = 1.27920350061
======= #Nearest Neighbors (kd_tree) with updated training file
- RMSLE nearest neighbors with neighbors=4 = 0.989827747648
======= RMSLE naiveBayes = 2.16326900532
=======
It is 1.25045194287.
1.26517399599
My score is the second lowest: 1.32651746013
My score for this is 1.13790775751
My score for this is 1.30120410415
Score is 1.09986335589
- L1 Penalty with C=10.0
- 2.60695415935
- L2 Penalty with C=10.0
- 2.6547980647
- L1 Penalty with C=100.0
- 2.57870120062
- L2 Penalty with C=100.0
- 2.585980683
- L1 Penalty with C=1000.0
- 2.53797970939
- L2 Penalty with C=1000.0
- 2.61280057305
- L1 Penalty with C=10000.0
- 2.56804227912
- L2 Penalty with C=10000.0
- 2.57051736226
- L1 Penalty with C=100000.0
- 2.55627986963
- L2 Penalty with C=100000.0
- 2.59808068079
- 2.70629841642