Questions tagged [train]

training (or estimation) of statistical models or algorithms.

Filter by
Sorted by
Tagged with
1
vote
1answer
18 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
6 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
15 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 ...
14
votes
5answers
2k 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
35 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
15 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
26 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 Matérn 5/2 and Matérn 5/2 ARD kernels. I notice that, with increasing ...
0
votes
0answers
46 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
22 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 ...
1
vote
1answer
110 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
17 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
15 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
10 views

Cross validation for timeseries interpolation

I am trying to calculate the MSE from interpolating few data points obtained from a sensor. I am considering 100 datapoints and applying 10 fold cross validation to it. For explaining what i am doing ...
0
votes
0answers
48 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
27 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
57 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
23 views

synthesized data to train classifier

Our dataset is relatively small (303 x 14) and so we decided to use synthpop package in R. The basic idea of synthetic data is to replace some or all of the observed values by sampling from ...
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
39 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
22 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
81 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 ...
0
votes
0answers
243 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 ...
0
votes
0answers
14 views

oversampling data with subclass

Oversampling of under-represented data is a way to combat class imbalance. For example, if we have a training data set with 100 data points of class A and 1000 data points of class B, we can over ...
0
votes
0answers
25 views

Regularized linear regression with class imbalance

I am trying to build a Linear Regression model using a not so big dataset. I'm more comfortable doing classification and I am not really an expert in regression. In classification, I was used to ...
0
votes
1answer
31 views

What to do with very little training data

What are some popular solutions in dealing with very little training data? Do these solutions rely on generating more data (e.g. bootstrapping, SMOTE, etc.)? Or do they rely on methods that do not ...
3
votes
1answer
131 views

Does retraining a model on all available data necessarily yield a better model?

A (simplified) typical workflow in machine learning might be: Train $m$ models on a training set. Validate the $m$ models on a validation set to yield the best model with parameters $\theta$. Retrain ...
0
votes
0answers
18 views

Splitting data into test and training when there are a low number of observations for a level within a variable

Context If a level within a categorical variable has less observations than there are elements in the training set then there's a chance that all of the elements of that level will be contained ...
0
votes
0answers
15 views

Different values in test / training data variable

The code and error in the following MWE represent my issue. code ...
0
votes
0answers
68 views

Credit Scoring WoE Calculation

I'm creating credit scoring model and stuck with WoE calculation. I know the formula and I know how to compute WoE for train sample. Should I use train sample WoE for test sample or I should compute ...
0
votes
1answer
43 views

What's this class of algorithms called: (entire training dataset, new input) -> output?

Supervised machine learning algorithms normally work by preprocessing a training dataset and outputting a compact model (e.g. a bunch of regression coefficients) that can quickly give an approximate ...
2
votes
1answer
47 views

Save Machine Learning Model progress for later [closed]

another dumb question, but how do you save the progress an ML model has made and start from that point later? Its kind of a vague question, but this is an example of what I am talking about: Say, ...
0
votes
0answers
19 views

ANN for Boundary Value Problem

I have a question regarding solving Boundary Value Problems (BVP) using ANNs. My understanding is that this is currently a challenging task. Most scientific literature on the subject is interested in ...
0
votes
1answer
17 views

With limited training time - should I shrink the training set and run more epochs?

I wonder what is considered a better practice in deep learning. I have a dataset of 100K images that I want to use for training a regional-CNN for a whole week. Is it possible that my network will be ...
4
votes
2answers
110 views

Practical realities of updating a trained model with new data

In my day to day work, I train models on data using R packages that have no extension for Bayesian priors. I will generally have a large dataset to start off with, and add new data as needed. Any ...
3
votes
0answers
320 views

How to train CNN for noise removal from images using Matlab [closed]

I am currently working on a project of mine where I want to use Convolutional neural networks for noise removal from images. I am talking about removing Poisson type of noise. The software that I am ...
2
votes
1answer
109 views

deep learning : how to retrain an already trained model on different data?

I have a model (a convolutional neural network) that has already been trained on set of training data for human pose estimation. It works very well for simple cases, but I want to make it more robust ...
1
vote
1answer
35 views

The performance metric used in prediction is different from the objective function to train the model

For linear regression and many machine learning models, we use the same performance metric during the training and testing stage. For example, during the training stage, our machine learning algorithm ...
0
votes
0answers
141 views

What is the training algorithm for sklearn's Bayesian Ridge Regression?

I read the sklearn's code to train Bayesian Ridge Regression, but can not understand the algorithm. I think it is EM, but don't know where the update equations for lambda_ and alpha_ come from. Any ...
0
votes
0answers
30 views

Does maximizing model generativeness maximize its discriminative-ness?

If I train a model M in a way that maximizes the penalized likelihood of M given some data, is this equivalent to maximizing its ...
0
votes
0answers
472 views

RFECV + GRIDSEARCHCV on entire dataset (?)

I have 30,000 samples with 150 features for a binary classification problem, now I plan to follow: https://stackoverflow.com/questions/23815938/recursive-feature-elimination-and-grid-search-using-...
0
votes
0answers
37 views

If I remove 20% of the train examples, my CV and train score improves, how can I find the reason why?

I have a small dataset - several thousands examples with a lot of features, a lot of them have a lot of zeros. If I remove 20% of examples from the end of the dataset - my CV and train(on changed ...
2
votes
1answer
141 views

How to meaningfully compute the accuracy of a multi-step forecast produced by a model

I am trying to measure the accuracy of my model in producing a multi-step forecast and I have read a lot of different opinions on the matter and am now rather confused. The goal of my model is to ...
2
votes
1answer
439 views

in-sample data vs out-of-sample data

I know that a train-validation-test splits the data into: a training dataset - obviously my "in-sample" data a validation dataset a test data set - obviously my "out-of-sample" data My question is: ...
3
votes
2answers
447 views

Can overfitting be a good thing in some cases?

I know the goal of machine learning is to create generalizable models and therefore overfitting is undesirable. However, I wonder if it could be desirable in some cases. For example, let's say I ...
0
votes
0answers
25 views

How to test a PCA+classifier model? [duplicate]

I have a 100x45 dataset and I'd like to perform feature selection and classification/regression. I'm currently using various techniques to check which one has the best performances, but I have a ...
0
votes
1answer
223 views

Why is gradient descent and it's variants used instead of BFGS and L-BFGS for training neural nets? [duplicate]

My understanding is that BFGS and L-BFGS solve the same type of optimization problems as GD and it's variants. Why is GD the go to algorithm for training neural networks?
2
votes
3answers
152 views

Data set splitting for statistical inference?

probably a very basic question -- I am modelling companies' decisions on which mode of payment they use in M&A deals with a help of logit model. I am so far interested only in what variables are ...
1
vote
0answers
545 views

Drawing ROC curves from RFE() training results in caret

I want to generate ROC curves using the training data and results from the rfe function in caret. I have managed to do this with the code below but there is some inconsistency between the ROC value ...