Tweet
Tweet

Slide 1

Slide 1 text

Better than My Meetup/Conference Talks: Going Deeper in Various GBM Topics Szilard Pafka, PhD Chief Scientist, Epoch (USA) GBM Advanced Workshop Budapest Nov 2019

Slide 2

Slide 2 text

Why GBMs

Slide 3

Slide 3 text

No content

Slide 4

Slide 4 text

meetup/conference talks going deeper section dividers

Slide 5

Slide 5 text

No content

Slide 6

Slide 6 text

Disclaimer: I am not representing my employer (Epoch) in this talk I cannot confirm nor deny if Epoch is using any of the methods, tools, results etc. mentioned in this talk

Slide 7

Slide 7 text

Source: Andrew Ng

Slide 8

Slide 8 text

Source: Andrew Ng

Slide 9

Slide 9 text

Source: Andrew Ng

Slide 10

Slide 10 text

No content

Slide 11

Slide 11 text

No content

Slide 12

Slide 12 text

No content

Slide 13

Slide 13 text

No content

Slide 14

Slide 14 text

No content

Slide 15

Slide 15 text

No content

Slide 16

Slide 16 text

...

Slide 17

Slide 17 text

No content

Slide 18

Slide 18 text

No content

Slide 19

Slide 19 text

No content

Slide 20

Slide 20 text

No content

Slide 21

Slide 21 text

No content

Slide 22

Slide 22 text

No content

Slide 23

Slide 23 text

No content

Slide 24

Slide 24 text

No content

Slide 25

Slide 25 text

http://lowrank.net/nikos/pubs/empirical.pdf http://www.cs.cornell.edu/~alexn/papers/empirical.icml06.pdf

Slide 26

Slide 26 text

http://lowrank.net/nikos/pubs/empirical.pdf http://www.cs.cornell.edu/~alexn/papers/empirical.icml06.pdf

Slide 27

Slide 27 text

No content

Slide 28

Slide 28 text

No content

Slide 29

Slide 29 text

No content

Slide 30

Slide 30 text

No content

Slide 31

Slide 31 text

top algos (RF, boosting), all features 2007

Slide 32

Slide 32 text

top algos (RF, boosting), all features most algos (lin, tree, nnet) worst algos (knn, NB) 2007

Slide 33

Slide 33 text

top algos (RF, boosting), all features most algos (lin, tree, nnet) worst algos (knn, NB) top algos, removed top feature(s) 2007

Slide 34

Slide 34 text

No content

Slide 35

Slide 35 text

No content

Slide 36

Slide 36 text

No content

Slide 37

Slide 37 text

Source: Hastie etal, ESL 2ed

Slide 38

Slide 38 text

Source: Hastie etal, ESL 2ed

Slide 39

Slide 39 text

GBM libs

Slide 40

Slide 40 text

No content

Slide 41

Slide 41 text

No content

Slide 42

Slide 42 text

No content

Slide 43

Slide 43 text

No content

Slide 44

Slide 44 text

10x

Slide 45

Slide 45 text

10x

Slide 46

Slide 46 text

No content

Slide 47

Slide 47 text

No content

Slide 48

Slide 48 text

No content

Slide 49

Slide 49 text

No content

Slide 50

Slide 50 text

No content

Slide 51

Slide 51 text

No content

Slide 52

Slide 52 text

No content

Slide 53

Slide 53 text

No content

Slide 54

Slide 54 text

No content

Slide 55

Slide 55 text

No content

Slide 56

Slide 56 text

No content

Slide 57

Slide 57 text

No content

Slide 58

Slide 58 text

No content

Slide 59

Slide 59 text

No content

Slide 60

Slide 60 text

No content

Slide 61

Slide 61 text

No content

Slide 62

Slide 62 text

No content

Slide 63

Slide 63 text

No content

Slide 64

Slide 64 text

No content

Slide 65

Slide 65 text

No content

Slide 66

Slide 66 text

No content

Slide 67

Slide 67 text

Scoring

Slide 68

Slide 68 text

No content

Slide 69

Slide 69 text

No content

Slide 70

Slide 70 text

No content

Slide 71

Slide 71 text

No content

Slide 72

Slide 72 text

No content

Slide 73

Slide 73 text

* very first request not shown >500ms (JVM “warmup”)

Slide 74

Slide 74 text

No content

Slide 75

Slide 75 text

No content

Slide 76

Slide 76 text

No content

Slide 77

Slide 77 text

No content

Slide 78

Slide 78 text

No content

Slide 79

Slide 79 text

No content

Slide 80

Slide 80 text

No content

Slide 81

Slide 81 text

No content

Slide 82

Slide 82 text

No content

Slide 83

Slide 83 text

No content

Slide 84

Slide 84 text

GBM-perf github repo

Slide 85

Slide 85 text

No content

Slide 86

Slide 86 text

No content

Slide 87

Slide 87 text

No content

Slide 88

Slide 88 text

No content

Slide 89

Slide 89 text

No content

Slide 90

Slide 90 text

No content

Slide 91

Slide 91 text

No content

Slide 92

Slide 92 text

multi-core/socket

Slide 93

Slide 93 text

No content

Slide 94

Slide 94 text

No content

Slide 95

Slide 95 text

No content

Slide 96

Slide 96 text

No content

Slide 97

Slide 97 text

No content

Slide 98

Slide 98 text

No content

Slide 99

Slide 99 text

CPU 1

Slide 100

Slide 100 text

CPU 1 CPU 2

Slide 101

Slide 101 text

CPU 1 CPU 2

Slide 102

Slide 102 text

CPU 1 CPU 2

Slide 103

Slide 103 text

CPU 1 CPU 2

Slide 104

Slide 104 text

No content

Slide 105

Slide 105 text

5x 3.5x

Slide 106

Slide 106 text

No content

Slide 107

Slide 107 text

No content

Slide 108

Slide 108 text

No content

Slide 109

Slide 109 text

No content

Slide 110

Slide 110 text

No content

Slide 111

Slide 111 text

No content

Slide 112

Slide 112 text

No content

Slide 113

Slide 113 text

No content

Slide 114

Slide 114 text

No content

Slide 115

Slide 115 text

No content

Slide 116

Slide 116 text

No content

Slide 117

Slide 117 text

No content

Slide 118

Slide 118 text

No content

Slide 119

Slide 119 text

No content

Slide 120

Slide 120 text

zero

Slide 121

Slide 121 text

No content

Slide 122

Slide 122 text

No content

Slide 123

Slide 123 text

Spark

Slide 124

Slide 124 text

No content

Slide 125

Slide 125 text

No content

Slide 126

Slide 126 text

No content

Slide 127

Slide 127 text

No content

Slide 128

Slide 128 text

No content

Slide 129

Slide 129 text

No content

Slide 130

Slide 130 text

No content

Slide 131

Slide 131 text

No content

Slide 132

Slide 132 text

No content

Slide 133

Slide 133 text

No content

Slide 134

Slide 134 text

No content

Slide 135

Slide 135 text

No content

Slide 136

Slide 136 text

No content

Slide 137

Slide 137 text

No content

Slide 138

Slide 138 text

No content

Slide 139

Slide 139 text

No content

Slide 140

Slide 140 text

No content

Slide 141

Slide 141 text

No content

Slide 142

Slide 142 text

No content

Slide 143

Slide 143 text

No content

Slide 144

Slide 144 text

No content

Slide 145

Slide 145 text

No content

Slide 146

Slide 146 text

No content

Slide 147

Slide 147 text

No content

Slide 148

Slide 148 text

No content

Slide 149

Slide 149 text

No content

Slide 150

Slide 150 text

No content

Slide 151

Slide 151 text

No content

Slide 152

Slide 152 text

GPU

Slide 153

Slide 153 text

No content

Slide 154

Slide 154 text

No content

Slide 155

Slide 155 text

No content

Slide 156

Slide 156 text

No content

Slide 157

Slide 157 text

catboost

Slide 158

Slide 158 text

No content

Slide 159

Slide 159 text

No content

Slide 160

Slide 160 text

No content

Slide 161

Slide 161 text

No content

Slide 162

Slide 162 text

No content

Slide 163

Slide 163 text

No content

Slide 164

Slide 164 text

No content

Slide 165

Slide 165 text

No content

Slide 166

Slide 166 text

No content

Slide 167

Slide 167 text

No content

Slide 168

Slide 168 text

API / tuning

Slide 169

Slide 169 text

No content

Slide 170

Slide 170 text

No content

Slide 171

Slide 171 text

No content

Slide 172

Slide 172 text

No content

Slide 173

Slide 173 text

No content

Slide 174

Slide 174 text

http://www.jmlr.org/papers/volume13/bergstra12a/bergstra12a.pdf

Slide 175

Slide 175 text

http://www.argmin.net/2016/06/20/hypertuning/

Slide 176

Slide 176 text

http://www.argmin.net/2016/06/20/hypertuning/

Slide 177

Slide 177 text

No content

Slide 178

Slide 178 text

No content

Slide 179

Slide 179 text

No content

Slide 180

Slide 180 text

time ordered data time ordered data

Slide 181

Slide 181 text

time ordered data time ordered data train sample

Slide 182

Slide 182 text

time ordered data time ordered data train test sample sample (slightly different distribution)

Slide 183

Slide 183 text

time ordered data time ordered data train test sample sample proper train early stopping Model selection resampled 80-10-10 (~CV) (slightly different distribution)

Slide 184

Slide 184 text

time ordered data time ordered data train test sample sample proper train early stopping Model selection random search over lightgbm resampled 80-10-10 (~CV) (slightly different distribution)

Slide 185

Slide 185 text

No content

Slide 186

Slide 186 text

No content

Slide 187

Slide 187 text

No content

Slide 188

Slide 188 text

No content

Slide 189

Slide 189 text

No content

Slide 190

Slide 190 text

No content

Slide 191

Slide 191 text

No content

Slide 192

Slide 192 text

No content

Slide 193

Slide 193 text

Closing

Slide 194

Slide 194 text

No content

Slide 195

Slide 195 text

No content

Slide 196

Slide 196 text

Source: https://www.linkedin.com/pulse/winning-solution-kaggledays-2019-competition-san-francisco-mark-peng/

Slide 197

Slide 197 text

Source: https://www.linkedin.com/pulse/winning-solution-kaggledays-2019-competition-san-francisco-mark-peng/

Slide 198

Slide 198 text

Source: https://www.linkedin.com/pulse/winning-solution-kaggledays-2019-competition-san-francisco-mark-peng/

Slide 199

Slide 199 text

Source: https://www.linkedin.com/pulse/winning-solution-kaggledays-2019-competition-san-francisco-mark-peng/

Slide 200

Slide 200 text

Source: https://www.linkedin.com/pulse/winning-solution-kaggledays-2019-competition-san-francisco-mark-peng/

Slide 201

Slide 201 text

Source: https://www.linkedin.com/pulse/winning-solution-kaggledays-2019-competition-san-francisco-mark-peng/

Slide 202

Slide 202 text

No content

Slide 203

Slide 203 text

More:

Slide 204

Slide 204 text

No content