Machine learning supervised regression problem on auto mpg.

Input data

	N	Mean	SD	Min	Max
cylinders	398	5.454774	1.7010042	3	8.0
displacement	398	193.425879	104.2698382	68	455.0
horsepower	392	104.469388	38.4911599	46	230.0
weight	398	2970.424623	846.8417742	1613	5140.0
acceleration	398	15.568091	2.7576889	8	24.8
model	398	76.010050	3.6976266	70	82.0
origin	398	1.572864	0.8020549	1	3.0

Variable for ml response (mpg)

	N	Mean	SD	Min	Max
class	398	23.51457	7.815984	9	46.6

ML task

target_names	num_col	num_row
class	8	398

Regression measures on input data

regr.mse	regr.rmse	regr.maxae
5.8795	2.424768	7.674978

Random sample of 5 response values

row_ids	truth	response
99	16	15.71012
386	38	34.62937
57	26	26.14580
230	16	15.33663
48	19	17.51291

Learner

task_type	regr
predict_type	response
scale.robust	0
regr.xgboost.alpha	0.00841273302842034
regr.xgboost.colsample_bylevel	0.870184205728583
regr.xgboost.colsample_bytree	0.786294631077908
regr.xgboost.eta	0.00464192861730337
regr.xgboost.lambda	0.00108953676875045
regr.xgboost.nrounds	3363
regr.xgboost.nthread	1
regr.xgboost.subsample	0.342429502122104
regr.xgboost.verbose	0

Timings

train	predict
0.681	0.044

New data for ml response

row_ids	truth	response
1	NA	17.97798

A more complete example of machine learning

leavesml with mlr3

• leavesml_mlr3.R

Notes classification measures

• regr.mse, Mean Squared Error
• regr.rmse, Root Mean Squared Error
• regr.maxae, Max Absolute Error

Notes random sample

• truth, target variable from the input data
• response, output from the machine learning algorithm

Notes data source

• Auto-Mpg Data

## Last Page Update Mon Aug 29 19:03:00 2022