Questions tagged [train]
training (or estimation) of statistical models or algorithms.
271
questions
8
votes
3answers
557 views
Is it in general helpful to add “external” datasets to the training dataset?
Several people have already asked "is more data helpful?":
What impact does increasing the training data have on the overall system accuracy?
Can increasing the amount of training data make ...
0
votes
0answers
25 views
How to evaluate xgb.cv and is it okay to use min(test_logloss) even if there is large gab between train_logloss?
Please migrate if this is not correct spot the post this question.
I have a dataset including 70K - 1s and 300K - 0s, a type of classifier. I am using Xgboost to ...
0
votes
0answers
13 views
To what extend can pre-training and training affect the results of a prediction model?
If we pre-train a model forecasting COVID-19 with data of SARS, which had a different transmission pattern, will our model be weakened?
If we train the same model with data of (for instance) the USA, ...
0
votes
1answer
33 views
Training of a deep Artificial Neural Network
I have few doubts related to training a neural network with more parameters (weights and biases) than number of data points. I know there exists discussion (on this platform) related to training such ...
0
votes
1answer
27 views
Why does shuffling in train test split have a big impact with my loss and accuracy?
I used Keras for the train test split.
This is what I get when I shuffle during my train test split:
When I disable shuffle by setting shuffle:False this is what I ...
0
votes
0answers
9 views
svm train score is 1.0 and test score 0.996. How to verify its correct?
My dataset has features CO2, PM2.5, Temperature and Humidity and I am trying to predict the Air Quality Index based on this information. I have approximately 24,000 data points for each of the ...
2
votes
1answer
22 views
Is it possible to overfitting within single epoch
Let me put my question first. For a time-series prediciton, is it possible to overfit even within the first epoch, when training data and validation data should all "new" to model?
Features and ...
0
votes
0answers
39 views
Machine Learning: Why do I have this pattern of train and validation accuracy?
I am trying to understand what would generate this pattern of accuracy in train and validation dataset (second and third plot below).
I am training a network to recognize 6 types of faces (they are ...
0
votes
0answers
55 views
Can the size of the training set of a neural network be smaller compared to other models?
If some of you knows what is the rule thumb one in ten rule this what I want to discuss.
Neural network can, nowadays, go deeper in performances. Did you find, according to your experiences, that you ...
1
vote
2answers
14 views
Issues with training on a sample of training set?
I am training an SVM on highly imbalanced data. I have rectified this issue and my ML pipeline works just fine. I have allocated 70% of my dataset for training, however this takes an infeasible amount ...
0
votes
0answers
10 views
How to improve neural network training against a large data set of points with varying magnitude
I am currently using TensorFlow and have simply been trying to train a neural network directly against a large continuous data set, e.g. $y = [0.014, 1.545, 10.232, 0.948, ...]$. The loss function in ...
3
votes
1answer
72 views
What is the relationship between mean squared error and classification error?
I've trained a network using a genetic algorithm and I have two possible fitness functions for my GA: MSE and CErr.
If I use MSE as my fitness function, over time MSE decreases and classification ...
1
vote
1answer
52 views
Keras: What is the meaning of batch_size for validation?
If I understand correctly that batch size is the number of samples used in the training of a NN before the gradient gets updated, then why do we need a specified batch_size for the validation sample?
...
0
votes
0answers
11 views
Can I stop training my neural network at this point?
When the validation error of my Neural Network that I am trying to train is slowly decreasing but not by much, is it okay to stop train the network at that point, or do I need to increase the training ...
0
votes
0answers
33 views
How to estimate the performance of Neural network
I have a feedforward neural network with two hidden layers built in keras.
let say I have 40 observations. I split the data into train (e.g., 35 observations) and test (e.g., 5 observations) sets. ...
1
vote
2answers
27 views
What are the best models/methods for training when the target is multidimensional?
I am not very familiar with all methods in Machine Learning. However, I know for example when I apply linear regression, the y is always assumed to be one-dimensional.
My target is multi-dimensional. ...
0
votes
0answers
8 views
Train a neural network as part of other ML algorithm
I want to implement some kind of neural network as "encoder" that encode my input, transform it into other dimension and use the output (encoded data) in other algorithm like kNN etc.
My question is, ...
0
votes
0answers
22 views
Increased computation time for training and prediction with reduced feature space?
I implemented a PCA algorithm to reduce the input feature space of my neural network from 230 to 110 features.
My naive expectation was that if I train a neural network using the same hyper ...
0
votes
0answers
37 views
Data leakage: Does it make sense to split time series this way?
Iām trying to make sense of how to create the training data for time series.
Letās say we have sales data:
...
1
vote
1answer
26 views
Cross validation in trainControl function
If we use cross validation in trainControl function, still do we need to perform the prediction on test set or training data in train function is sufficient?
I split the data in training and testing,...
1
vote
0answers
41 views
is training dataset of machine learning are Big Data processing results? [closed]
I need to understand the application of machine learning in big data processing. I am so confused with the concept of big data processing and machine learning and I need clear responses. In general, ...
0
votes
0answers
20 views
Model loss stays the same for hours before dropping
I'm training a CNN to colorize images. The model I have is not incredibly deep, and should work fine on the card I'm training on (2080 TI). Initially, I suspected the model was flawed in some way ...
2
votes
0answers
18 views
Goodness of Fit Test vs Testing RMSE
I have a bit of a broad question. It seems to be that there are two different approaches (Borrowing a bit of Breiman's Two Culture's paper) when it comes to testing if our data looks like our model.
...
0
votes
0answers
22 views
Modeling when unlimited data generator is available
For simplicity, a regression task is needed to be done to model the inverse function of f(x). Let us say:
...
0
votes
0answers
49 views
Oversampling/Undersampling in respect to Train and Test - Isolation Forest
I've got a quite imbalanced data set.
144.496 : 162 -> ratio of 1000:1
I would like to use IsolationForest to detect the 162 anomalys.
I've already split the data.
However, the iForest doesn't ...
0
votes
1answer
26 views
Validation loss is decreasing, accuracy is decreasing too
So, I have the following charts from my experience.Can any one explain why accuracy is decreasing while the loss in train and validation is decreasing? The point is that i can't early stop too in the ...
0
votes
0answers
38 views
object detection loss
I have trained an ssd detector in my own dataset and the values of train loss and val loss are shown in the picture. However in all the epochs the value of val loss is lower than that of train loss ? ...
0
votes
0answers
17 views
I have capped my response variable, should I calculate my RMSE/MAE/MAPE with the true values capped or not?
So, I have trained a model in my train set with the response variable with a superior limit. Because, the peaks are not important for my analysis. And if I dropped it, I would lost a lot of data.
...
2
votes
1answer
24 views
Should a model be trained until it is stable to find optimal hyperparameters?
A model may take several days to train until it reaches an equilibrium - say if the change in error between epochs is lower than some threshold $\epsilon$, or accuracy reaches some equilibrium.
When ...
0
votes
0answers
15 views
Is it possible to achieve both stratified sampling and keeping the same train/test dataset split across different runs?
Generally, it is suggested to sample a dataset such that test set and train set remain the same when running the code multiple times, for comparison but also to hide your algorithm the whole dataset. ...
1
vote
1answer
31 views
should I re-initialize my optimizer and my scheduler before I try to fine tune my neural network on the different dataset?
I am doing NLP, and I have this block of Transformer body that was already trained on dataset A.
Now I am interested in fine tuning this same Transformer on a new dataset B.
In my Python code, should ...
15
votes
5answers
3k views
Can increasing the amount of training data make overfitting worse?
Suppose I train a neural network on dataset A and evaluate on dataset B (that has a different feature distribution than dataset A). If I increase the amount of data in dataset A by a factor of 10, is ...
3
votes
1answer
71 views
About the need of splitting data in stacking
I learned stacking of machine learning in a book, hands-on machine learning 2nd edition (2019).
The picture was cited from hands-on machine learning 2nd edition (2019).
In the above situation, ...
0
votes
0answers
45 views
Aggregate data for machine learning. Weights or fake disaggregation?
I have a dataset of medical centers and I need to predict their infection rate, based on the center characteristics and aggregated patient data (eg. percentage of patients which underwent a certain ...
0
votes
0answers
58 views
Gaussian process with ARD kernel much more expensive to train
I'm fitting a Gaussian process regression model in MATLAB (using the quasi-Newton method) with 10 input parameters, using the MatƩrn 5/2 and MatƩrn 5/2 ARD kernels.
I notice that, with increasing ...
0
votes
0answers
102 views
Why backpropagation if loss function is not convex in nature?
Backpropagation contains the method of gradient decent, which works well for convex loss functions with a global minima.
But, for training, in most of the neural network tasks, backpropagation is ...
1
vote
1answer
74 views
Test score bigger than Train score in Linear Regression
I'm new to ML and I'm trying to create a linear regression model. My data consist of 100 samples with 4 features each. This is my humble code
...
2
votes
1answer
441 views
Why my validation accuracy and AUC are higher than my training accuracy and AUC?
I have a binary classification problem and I use LightGBM classifier to build my model based on 5 features. I divided my dataset (94 observations) into two parts:
Training dataset: 60 observations
...
1
vote
1answer
31 views
Can we apply SMOTE on data with k-fold CV
The SMOTE for the imbalance should be applied for the training data only, right? Can we still do it (perform SMOTE on training data) while we select the k-fold CV and does not go for splitting the ...
0
votes
1answer
18 views
Removing parameters from the training set helped training
I have a large dataset with 10s of millions of points in a 10 dimensional parameter space. I have tried training my regression neural network on the entire parameter space and got decent (ish) results....
0
votes
0answers
50 views
Specificity decreasing when new features are added to glmnet model for case/control prediction
I'm using glmnet for prediction of case/control, which I created with the function train with additional parameters for cross ...
0
votes
0answers
18 views
Training a deep neural network using a changing (increasing) dataset
I am trying to train a deep autoencoder (applies also for other architectures) in the following way:
Step 1) I start with a fix dataset of e.g. 10k samples.
Step 2) A training "loop" consists of ...
1
vote
0answers
55 views
how to select training and testing data for interpolation in 100 instances data?
I would like to divide my data of only 100 instances into training and testing an use the training data to fit a curve(interpolate) and use the testing data to calculate the error at the interpolated ...
1
vote
1answer
60 views
Logistic regression - what is being predicted? [closed]
R has more than one way to create logistic regressions to predict binary outcomes.
Here's the code that I'm using that is giving strange answers.
...
0
votes
1answer
20 views
Randomize dataset for Restricted Boltmann Machines
Suppose I want to train a RBM (or even a DBN architecture) and then fine-tune the parameter training a Feedforward NN. In my case the dataset is composed of time series, so in principle there is a ...
0
votes
0answers
13 views
Difference between retraining on different portions of data and training initially on larger data set
I have a large data set that doesn't fit in memory and would have to use something like Keras's model.fit_generator if I would like to train the model on all of the ...
1
vote
1answer
159 views
Determining suitable seed number in R to randomly split data into training and test sets
Good day. I have sample size of 160 and I am randomly splitting them into 70% train and 30% test data sets. My question is about the set.seed() value which returns different random samples in each ...
2
votes
0answers
26 views
How do I check my GAN implementation is correct?
I wrote a GAN implementation and I trained that to produce some sample images after training on a dataset. The images looked visually fine.
Now I want to test my implementation on the CI and make ...
3
votes
1answer
213 views
Data leakage when using walk forward optimization
I am setting up a neural network that will predict the incoming customers at a store for the next seven days (the output is a list with seven numbers, one for each day). As input, I will give the ...
1
vote
0answers
315 views
Very High Training Accuracy and very low Testing Accuracy CNN
I'm using 3 layer CNN with 8, 16, and 32 filters, each of size 5 X 5. I'm getting an training accuracy of 99.97%. Testing accuracy of 41.11%.
Total classes: 605
Train Set: Each class has 7 samples
...
