I try to run code from this web site in my computer.
I use data set from kaggle competition
In my training data 1022 rows and 81 variables.
I run this code:
hyper_grid <- expand.grid(
shrinkage = c(.01, .1, .3),
interaction.depth = c(1, 3, 5),
n.minobsinnode = c(5, 10, 15),
bag.fraction = c(.65, .8, 1),
optimal_trees = 0, # a place to dump results
min_RMSE = 0 # a place to dump results
)
random_index <- sample(1:nrow(ames_train), nrow(ames_train))
random_ames_train <- ames_train[random_index, ]
# grid search
for(i in 1:nrow(hyper_grid)) {
# reproducibility
set.seed(123)
# train model
gbm.tune <- gbm(
formula = SalePrice ~ .,
distribution = "gaussian",
data = random_ames_train,
n.trees = 5000,
interaction.depth = hyper_grid$interaction.depth[i],
shrinkage = hyper_grid$shrinkage[i],
n.minobsinnode = hyper_grid$n.minobsinnode[i],
bag.fraction = hyper_grid$bag.fraction[i],
train.fraction = .75,
n.cores = NULL, # will use all cores by default
verbose = FALSE
)
I'm waiting more than 1 hour.
I think it's bacause my laptop is not powerful.
On the picture you can see parameters of my computer.
Please, answer: can my computer perform this operation?
If yes, how long should I wait?
It's taking a long time because you're training 81 GBM models, and GBM's are complex. To get a rough estimate of training time, you could train one model and then multiply that time by 81.
I am trying to train a neural network using train function and neuralnet as my method paramater to predict times table.
I am scaling my training data set as well.
Even though I've tried different learningrates, stepmaxes, and thresholds for my neuralnet, each time I tried to train the network using train function one of the k-folds happened to fail every time saying
1: Algorithm did not converge in 1 of 1 repetition(s) within the stepmax.
2: predictions failed for Fold05.Rep1: layer1=8, layer2=0, layer3=0 Error in cbind(1, pred) %*% weights[[num_hidden_layers + 1]] :
requires numeric/complex matrix/vector arguments
I am guessing that this is because of weights being random so somehow each time I happen to get some weights that are not going to converge.
Is there anyway of preventing this? Maybe trying to re-train the particular fold which has failed using different weights?
Here is my code:
library(caret)
library(neuralnet)
# Create the dataset
tt = data.frame(multiplier = rep(1:10, times = 10), multiplicand = rep(1:10, each = 10))
tt = cbind(tt, data.frame(product = tt$multiplier * tt$multiplicand))
# Splitting
indexes = createDataPartition(tt$product,
times = 1,
p = 0.7,
list = FALSE)
tt.train = tt[indexes,]
tt.test = tt[-indexes,]
# Pre-process
preProc <- preProcess(tt, method = c('center', 'scale'))
tt.preProcessed <- predict(preProc, tt)
tt.preProcessed.train <- tt.preProcessed[indexes,]
tt.preProcessed.test <- tt.preProcessed[-indexes,]
# Train
train.control <- trainControl(method = "repeatedcv",
number = 10,
repeats = 3)
tune.grid <- expand.grid(layer1 = 8,
layer2 = 0,
layer3 = 0)
tt.cv <- train(product ~ .,
data = tt.preProcessed.train,
method = 'neuralnet',
tuneGrid = tune.grid,
trControl = train.control,
linear.output = TRUE,
algorithm = 'backprop',
learningrate = 0.01,
stepmax = 500000,
lifesign = 'minimal',
threshold = 0.01)
I've been exploring the xgboost package in R and went through several demos as well as tutorials but this still confuses me: after using xgb.cv to do cross validation, how does the optimal parameters get passed to xgb.train? Or should I calculate the ideal parameters (such as nround, max.depth) based on the output of xgb.cv?
param <- list("objective" = "multi:softprob",
"eval_metric" = "mlogloss",
"num_class" = 12)
cv.nround <- 11
cv.nfold <- 5
mdcv <-xgb.cv(data=dtrain,params = param,nthread=6,nfold = cv.nfold,nrounds = cv.nround,verbose = T)
md <-xgb.train(data=dtrain,params = param,nround = 80,watchlist = list(train=dtrain,test=dtest),nthread=6)
Looks like you misunderstood xgb.cv, it is not a parameter searching function. It does k-folds cross validation, nothing more.
In your code, it does not change the value of param.
To find best parameters in R's XGBoost, there are some methods. These are 2 methods,
(1) Use mlr package, http://mlr-org.github.io/mlr-tutorial/release/html/
There is a XGBoost + mlr example code in the Kaggle's Prudential challenge,
But that code is for regression, not classification. As far as I know, there is no mlogloss metric yet in mlr package, so you must code the mlogloss measurement from scratch by yourself. CMIIW.
(2) Second method, by manually setting the parameters then repeat, example,
param <- list(objective = "multi:softprob",
eval_metric = "mlogloss",
num_class = 12,
max_depth = 8,
eta = 0.05,
gamma = 0.01,
subsample = 0.9,
colsample_bytree = 0.8,
min_child_weight = 4,
max_delta_step = 1
)
cv.nround = 1000
cv.nfold = 5
mdcv <- xgb.cv(data=dtrain, params = param, nthread=6,
nfold=cv.nfold, nrounds=cv.nround,
verbose = T)
Then, you find the best (minimum) mlogloss,
min_logloss = min(mdcv[, test.mlogloss.mean])
min_logloss_index = which.min(mdcv[, test.mlogloss.mean])
min_logloss is the minimum value of mlogloss, while min_logloss_index is the index (round).
You must repeat the process above several times, each time change the parameters manually (mlr does the repeat for you). Until finally you get best global minimum min_logloss.
Note: You can do it in a loop of 100 or 200 iterations, in which for each iteration you set the parameters value randomly. This way, you must save the best [parameters_list, min_logloss, min_logloss_index] in variables or in a file.
Note: better to set random seed by set.seed() for reproducible result. Different random seed yields different result. So, you must save [parameters_list, min_logloss, min_logloss_index, seednumber] in the variables or file.
Say that finally you get 3 results in 3 iterations/repeats:
min_logloss = 2.1457, min_logloss_index = 840
min_logloss = 2.2293, min_logloss_index = 920
min_logloss = 1.9745, min_logloss_index = 780
Then you must use the third parameters (it has global minimum min_logloss of 1.9745). Your best index (nrounds) is 780.
Once you get best parameters, use it in the training,
# best_param is global best param with minimum min_logloss
# best_min_logloss_index is the global minimum logloss index
nround = 780
md <- xgb.train(data=dtrain, params=best_param, nrounds=nround, nthread=6)
I don't think you need watchlist in the training, because you have done the cross validation. But if you still want to use watchlist, it is just okay.
Even better you can use early stopping in xgb.cv.
mdcv <- xgb.cv(data=dtrain, params=param, nthread=6,
nfold=cv.nfold, nrounds=cv.nround,
verbose = T, early.stop.round=8, maximize=FALSE)
With this code, when mlogloss value is not decreasing in 8 steps, the xgb.cv will stop. You can save time. You must set maximize to FALSE, because you expect minimum mlogloss.
Here is an example code, with 100 iterations loop, and random chosen parameters.
best_param = list()
best_seednumber = 1234
best_logloss = Inf
best_logloss_index = 0
for (iter in 1:100) {
param <- list(objective = "multi:softprob",
eval_metric = "mlogloss",
num_class = 12,
max_depth = sample(6:10, 1),
eta = runif(1, .01, .3),
gamma = runif(1, 0.0, 0.2),
subsample = runif(1, .6, .9),
colsample_bytree = runif(1, .5, .8),
min_child_weight = sample(1:40, 1),
max_delta_step = sample(1:10, 1)
)
cv.nround = 1000
cv.nfold = 5
seed.number = sample.int(10000, 1)[[1]]
set.seed(seed.number)
mdcv <- xgb.cv(data=dtrain, params = param, nthread=6,
nfold=cv.nfold, nrounds=cv.nround,
verbose = T, early.stop.round=8, maximize=FALSE)
min_logloss = min(mdcv[, test.mlogloss.mean])
min_logloss_index = which.min(mdcv[, test.mlogloss.mean])
if (min_logloss < best_logloss) {
best_logloss = min_logloss
best_logloss_index = min_logloss_index
best_seednumber = seed.number
best_param = param
}
}
nround = best_logloss_index
set.seed(best_seednumber)
md <- xgb.train(data=dtrain, params=best_param, nrounds=nround, nthread=6)
With this code, you run cross validation 100 times, each time with random parameters. Then you get best parameter set, that is in the iteration with minimum min_logloss.
Increase the value of early.stop.round in case you find out that it's too small (too early stopping). You need also to change the random parameter values' limit based on your data characteristics.
And, for 100 or 200 iterations, I think you want to change verbose to FALSE.
Side note: That is example of random method, you can adjust it e.g. by Bayesian optimization for better method. If you have Python version of XGBoost, there is a good hyperparameter script for XGBoost, https://github.com/mpearmain/BayesBoost to search for best parameters set using Bayesian optimization.
Edit: I want to add 3rd manual method, posted by "Davut Polat" a Kaggle master, in the Kaggle forum.
Edit: If you know Python and sklearn, you can also use GridSearchCV along with xgboost.XGBClassifier or xgboost.XGBRegressor
This is a good question and great reply from silo with lots of details! I found it very helpful for someone new to xgboost like me. Thank you. The method to randomize and compared to boundary is very inspiring. Good to use and good to know. Now in 2018 some slight revise are needed, for example, early.stop.round should be early_stopping_rounds. The output mdcv is organized slightly differently:
min_rmse_index <- mdcv$best_iteration
min_rmse <- mdcv$evaluation_log[min_rmse_index]$test_rmse_mean
And depends on the application (linear, logistic,etc...), the objective, eval_metric and parameters shall be adjusted accordingly.
For the convenience of anyone who is running a regression, here is the slightly adjusted version of code (most are the same as above).
library(xgboost)
# Matrix for xgb: dtrain and dtest, "label" is the dependent variable
dtrain <- xgb.DMatrix(X_train, label = Y_train)
dtest <- xgb.DMatrix(X_test, label = Y_test)
best_param <- list()
best_seednumber <- 1234
best_rmse <- Inf
best_rmse_index <- 0
set.seed(123)
for (iter in 1:100) {
param <- list(objective = "reg:linear",
eval_metric = "rmse",
max_depth = sample(6:10, 1),
eta = runif(1, .01, .3), # Learning rate, default: 0.3
subsample = runif(1, .6, .9),
colsample_bytree = runif(1, .5, .8),
min_child_weight = sample(1:40, 1),
max_delta_step = sample(1:10, 1)
)
cv.nround <- 1000
cv.nfold <- 5 # 5-fold cross-validation
seed.number <- sample.int(10000, 1) # set seed for the cv
set.seed(seed.number)
mdcv <- xgb.cv(data = dtrain, params = param,
nfold = cv.nfold, nrounds = cv.nround,
verbose = F, early_stopping_rounds = 8, maximize = FALSE)
min_rmse_index <- mdcv$best_iteration
min_rmse <- mdcv$evaluation_log[min_rmse_index]$test_rmse_mean
if (min_rmse < best_rmse) {
best_rmse <- min_rmse
best_rmse_index <- min_rmse_index
best_seednumber <- seed.number
best_param <- param
}
}
# The best index (min_rmse_index) is the best "nround" in the model
nround = best_rmse_index
set.seed(best_seednumber)
xg_mod <- xgboost(data = dtest, params = best_param, nround = nround, verbose = F)
# Check error in testing data
yhat_xg <- predict(xg_mod, dtest)
(MSE_xgb <- mean((yhat_xg - Y_test)^2))
I found silo's answer is very helpful.
In addition to his approach of random research, you may want to use Bayesian optimization to facilitate the process of hyperparameter search, e.g. rBayesianOptimization library.
The following is my code with rbayesianoptimization library.
cv_folds <- KFold(dataFTR$isPreIctalTrain, nfolds = 5, stratified = FALSE, seed = seedNum)
xgb_cv_bayes <- function(nround,max.depth, min_child_weight, subsample,eta,gamma,colsample_bytree,max_delta_step) {
param<-list(booster = "gbtree",
max_depth = max.depth,
min_child_weight = min_child_weight,
eta=eta,gamma=gamma,
subsample = subsample, colsample_bytree = colsample_bytree,
max_delta_step=max_delta_step,
lambda = 1, alpha = 0,
objective = "binary:logistic",
eval_metric = "auc")
cv <- xgb.cv(params = param, data = dtrain, folds = cv_folds,nrounds = 1000,early_stopping_rounds = 10, maximize = TRUE, verbose = verbose)
list(Score = cv$evaluation_log$test_auc_mean[cv$best_iteration],
Pred=cv$best_iteration)
# we don't need cross-validation prediction and we need the number of rounds.
# a workaround is to pass the number of rounds(best_iteration) to the Pred, which is a default parameter in the rbayesianoptimization library.
}
OPT_Res <- BayesianOptimization(xgb_cv_bayes,
bounds = list(max.depth =c(3L, 10L),min_child_weight = c(1L, 40L),
subsample = c(0.6, 0.9),
eta=c(0.01,0.3),gamma = c(0.0, 0.2),
colsample_bytree=c(0.5,0.8),max_delta_step=c(1L,10L)),
init_grid_dt = NULL, init_points = 10, n_iter = 10,
acq = "ucb", kappa = 2.576, eps = 0.0,
verbose = verbose)
best_param <- list(
booster = "gbtree",
eval.metric = "auc",
objective = "binary:logistic",
max_depth = OPT_Res$Best_Par["max.depth"],
eta = OPT_Res$Best_Par["eta"],
gamma = OPT_Res$Best_Par["gamma"],
subsample = OPT_Res$Best_Par["subsample"],
colsample_bytree = OPT_Res$Best_Par["colsample_bytree"],
min_child_weight = OPT_Res$Best_Par["min_child_weight"],
max_delta_step = OPT_Res$Best_Par["max_delta_step"])
# number of rounds should be tuned using CV
#https://www.hackerearth.com/practice/machine-learning/machine-learning-algorithms/beginners-tutorial-on-xgboost-parameter-tuning-r/tutorial/
# However, nrounds can not be directly derivied from the bayesianoptimization function
# Here, OPT_Res$Pred, which was supposed to be used for cross-validation, is used to record the number of rounds
nrounds=OPT_Res$Pred[[which.max(OPT_Res$History$Value)]]
xgb_model <- xgb.train (params = best_param, data = dtrain, nrounds = nrounds)
I am RSNNS to make a model. I am using QuickProp algorithm. here's my neural network:
mydata1 <- read.csv("-1-5_rand1.csv");
mydata <- mydata1[1:151, ]
test_set <- mydata1[152:168, ]
test_set1 <- test_set[c(-7)]
a <- SnnsRObjectFactory()
input <- mydata[c(-7)]
output <- mydata[c(7)]
b <- splitForTrainingAndTest(input, output, ratio = 0.22)
a <- mlp(b$inputsTrain, b$targetsTrain, size = 9, maxit = 650, learnFunc = "Quickprop", learnFuncParams = c(0.01, 2.5, 0.0001, 0, 0), updateFunc = "Topological_Order",
updateFuncParams = c(0.0), hiddenActFunc = "Act_TanH", computeError=TRUE, initFunc = "Randomize_Weights", initFuncParams = c(-1,1),
shufflePatterns = TRUE, linOut = FALSE, inputsTest = b$inputsTest, targetsTest = b$targetsTest)
I am predicting using test set as:
predictions <- predict(a, test_set1)
Is it possible to in RSNNS to predict after every 50 cycles using test set instead of predicting after 650 cycles?
the answer is you can't do it with the high-level interface, but with the low-level interface, you can have a look, e.g., at the mlp_irisSnnsR.R demo that is included in RSNNS
Hi when I am using the caret ensemble package I keep encountering this error that createTimeslices cross validation method cannot be used for caretEnsemble package.
Has anybody suggestion how to overcome this
I have solved the problem, and I am giving a reproducible example
library(quantmod)
startDate = as.Date("2010-01-01")
endDate = as.Date("2014-12-31")
getSymbols("^GDAXI", src = "yahoo", from = startDate, to = endDate)
RSI3<-RSI(Op(GDAXI), n= 3)
#Calculate a 3-period relative strength index (RSI) off the open price
EMA5<-EMA(Op(GDAXI),n=5)
#Calculate a 5-period exponential moving average (EMA)
EMAcross<- Op(GDAXI)-EMA5
#Let’s explore the difference between the open price and our 5-period EMA
MACD<-MACD(Op(GDAXI),fast = 12, slow = 26, signal = 9)
#Calculate a MACD with standard parameters
MACDsignal<-MACD[,2]
#Grab just the signal line to use as our indicator.
SMI<-SMI(Op(GDAXI),n=13,slow=25,fast=2,signal=9)
#Stochastic Oscillator with standard parameters
SMI<-SMI[,1]
#Grab just the oscillator to use as our indicator
PriceChange<- Cl(GDAXI) - Op(GDAXI)
#Calculate the difference between the close price and open price
Class<-ifelse(PriceChange>0,"UP","DOWN")
#Create a binary classification variable, the variable we are trying to predict.
DataSet<-data.frame(RSI3,EMAcross,MACDsignal,SMI,Class)
Create our data set
colnames(DataSet)<-c("RSI3","EMAcross","MACDsignal","Stochastic","Class")
#Name the columns
DataSet<-DataSet[-c(1:33),]
#Get rid of the data where the indicators are being calculated
Alldata<-cbind(DataSet,CombDF[34:1279,2])
colnames(Alldata)<-c("RSI3","EMAcross","MACDsignal","Stochastic","Class","ArabSpring")
TrainingSet<-Alldata[1:1000,]
TestSet<-Alldata[1001:1246,]
time_slices <- createTimeSlices(1:nrow(TrainingSet),initialWindow =800,horizon =200, fixedWindow = TRUE)
str(time_slices)
myTimeControl <- trainControl(method = "cv", number = 2, repeats = 1, savePrediction = TRUE,classProbs = TRUE,returnResamp = "final",returnData = TRUE,index= time_slices$train, indexOut=time_slices$test )
model_list_big <- caretList(
Class~., data=TrainingSet,
trControl=myTimeControl,
metric='Accuracy',
methodList=c('rf', 'gbm','treebag', 'nodeHarvest'),
tuneList=list(
rf=caretModelSpec(method='rf',tunelength = 10, ntrees = 2000, importance = TRUE),
gbm=caretModelSpec(method='gbm',tuneGrid= expand.grid(.interaction.depth = seq(1, 7, by = 2), .n.trees = seq(100,1000, by = 50), .shrinkage = c(0.01, 0.1) ) ),
tbag=caretModelSpec(method='treebag'),
Nharvest=caretModelSpec(method='nodeHarvest',nodes = 100)
)
)
greedy_ensemble <- caretEnsemble(model_list_big)
summary(greedy_ensemble)
ens_preds <- predict(greedy_ensemble, newdata=TestSet)