Questions tagged [train]
Training (or estimation) of statistical models or machine learning algorithms.
322
questions
0
votes
0answers
34 views
Is this K-Fold Cross-validation approach correct?
I've seen that Train-Validation-Test set technique is discussed and there is no consensus. Some people don't differ validation from test.
When I was studying this technique and K-fold cross-validation,...
0
votes
0answers
10 views
Continuation of Training for Neural Networks
originally posted in SO Artificial Intelligence, advised to post here instead
Background
I am developing a GAN model (based off this paper), and am trying to learn how to feed subsequent time series ...
0
votes
0answers
21 views
How does test_size relate when used in python sklearn for a 10 fold cross validation [closed]
I am trying to implement a ML algorithm in which I would like to use a 10 fold cross validation process but I would just like to get confirmation if my procedure is correct.
I am doing a binary ...
0
votes
0answers
21 views
Stratified sampling - use of proxy variable
For splitting of the data into train/test/val I use stratified sampling. Is it appropriate to define strata using information extracted from the dataset? E.g. use machine-learning to model proxy ...
2
votes
1answer
25 views
Time Series Forecasting Process With Regard to Training and Test Sets
I'm a bit confused about the process order in doing proper time series analysis/forecasting. Is it:
Stationary/seasonality checks, do any transformations required
Candidate model selection using ACF, ...
0
votes
0answers
14 views
Recurrent neural networks with loss of data
I want to train a recurrent neural network (RNN) for making predictions of some data.
I have 8 variable inputs and with them I have to make predictions of other two variables (outputs). I need RNN to ...
0
votes
0answers
10 views
How does glmnet in caret choose the values of lambda and how does it compute coefficients of the model?
I have a question that I've been struggling with. My students are asking me, but I can't figure it out myself.
When I train LASSO regression in R caret, I use the method "glmnet" and a grid ...
1
vote
0answers
10 views
Training in steps has any importance?
I'm trying to train a Siamese network for face Verification and eventually I came across the Contrastive Loss method for embedding vector distancing (kinda... I ...
0
votes
0answers
10 views
How do you use a weight matrix used for a batch in a final model to only take a single input?
Setup:
An input vector x to a neural network has 5 components, xi, i=1,5.
Say a weight vector is multiplied by x, w = [wi], i=1,5: so dot(w,x) = sum_i (wi * xi)
Now when doing batch training, say 10 ...
0
votes
0answers
5 views
Validation Loss gets better after adding Augmentation Layers but Test accuracy gets worse
I'm building a Siamese Network which should learn a face comparison function.
My model consists a CNN (which gets 2 inputs, and yields 2 embedding vectors). With the outputs I calculate:
...
0
votes
0answers
10 views
The implication for using only data points with extremely positive and negative labels for classification
Here I use the Quora dataset from Kaggle to explain my question. The label column of this dataset is is_duplicate and has value of 0 or 1. 255,043 question pairs in the dataset are labeled as 0(not ...
1
vote
0answers
20 views
How to improve model generalization with small dataset?
I have a fairly small dataset (45 data points (i.e populations) where I took plant information. I'm running a random forest regression on my measurements and climate information to predict a feature, ...
0
votes
1answer
71 views
About Epochs and how many of it?
I'm pritty new to the machine learning world, and I ws trying to figure out how many epochs should I run my training CNN model on the MNIST dataset (which has 60,000 training images and 10,000 ...
0
votes
1answer
42 views
Loss values above 1.0
I have a convolutional neural network for tensors classification in Pytorch. I am using Cross-Entropy Loss. My optimizer is Stochastic Gradient Descent and the learning rate is 0.0001. The accuracy of ...
0
votes
1answer
44 views
Train/Test split in multiple time series data
Currently, I have log files from 10 machines. Each log file contains an event type and the occurrence time of the event.
Each machine's log file is recorded under the same time frame, and I wish to ...
2
votes
2answers
135 views
Why validation accuracy is increasing very slowly?
My convolutional network seems to work well in learning the features. However, the accuracy of the validation set is increasing very slowly with respect to the learning rate as also illustrated in the ...
0
votes
0answers
16 views
When training data is much less than the prediction perdio
Given training data on tweets and their retweets, how would you predict the number of retweets of a given tweet after 7 days after only observing 2 days worth of data?
Its strikes me that this ...
2
votes
1answer
15 views
Does training time of CNN includes the validation time performed after every epoch?
I want to record train time of a CNN architecture in keras. My question is that while we are training a CNN, we also validate the model after every epoch to monitor that how well it generalizes which ...
0
votes
0answers
13 views
Correct way to use cross-validation for hyperparameter selection when testing a model in multiple trials which use different train-test split
I'm going to evaluate a model with some benchmark datasets. I want to perform 100 trials of training and testing of the model for each dataset, and I want to randomly split the data again in training ...
2
votes
1answer
276 views
How to choose a batch size and the number of epochs while training a NN
After searching I read diferent theorys that using a greater batch size has better performance while model is training, but in the other hand, I also find the oposite view, that using a mini-batch ...
3
votes
2answers
98 views
MultiClass Classification - Training OvO and OvA
I like to know how OvO (One vs One) and OvA (One vs All) models are trained in multiclass classification problem. To keep it simple, we have 4 classes, each of which has 1000 datapoints. What are the ...
0
votes
2answers
81 views
(How) can model parameters be learnt using MCMC?
I get stuck by part (b) of figure 4 in this paper: Hands-on Bayesian Neural Networks - a Tutorial for Deep Learning Users.
In my understanding, inference algorithms like MCMC are not for training ...
0
votes
0answers
21 views
Why is the Kernel Trick Matrix Symmetric Even When the Training and Test Set Differ in Size?
Ridge Regression can be expressed as $$\hat{y} = (\mathbf{X'X} + a\mathbf{I}_d)^{-1}\mathbf{X}x$$ where $\hat{y}$ is the predicted label, $\mathbf{I}_d$ the $d \times d$ identify matrix, $\mathbf{x}$ ...
0
votes
0answers
27 views
Random forest outside range of training - Alternatives?
Statistical models in general are not well suited for applications outside the training range.
Using a random forest binary classifier aimed at detecting an event, I've trained it using 100 data ...
15
votes
2answers
2k views
Can I (justifiably) train a second model only on the observations that a previous model predicted poorly?
Say I commit the following sins while building a predictive model:
I take my dataset and split it into four subsets: Three for training
(Train_A, Train_B, and Train_C) and one for validation.
I ...
0
votes
0answers
6 views
What is the training process of 1 input - 2 output network based on one mixed loss?
Let's say I have a neural network with one input, 2 output layers, 1 hidden layer. I define a combined loss based on the first and second output. For simplicity, assume that the first loss is RMSE and ...
1
vote
0answers
21 views
Training Individual-Level Predictor using Distribution of Group-level Data
I have a problem in which I'm looking to train an individual-level predictor for outcomes. I have information on individual-level covariates, but I don't have individual-level outcome variables. ...
1
vote
2answers
33 views
What are best practices around combining train/val/test splits when training a production model?
Say I'm training a deep neural network and I have split my data into train, val, and test splits. I have trained many models on the train data and then using the val data during the training loop for ...
3
votes
2answers
135 views
Split data into test, training and validation when some patients have multiple observations
I have a dataset that has multiple observations for some patients. Patients are labelled by patient IDs and I would like to split the data into testing, training and validating groups as I am trying ...
0
votes
0answers
37 views
How to force a neural network to uniformly decrease MAPE?
I aim for replicating an numerical (non stochastic) algorithm by a neural network. Therefore I have basically an unlimited amount of data and I wish that the network have an almost perfect fit in ...
1
vote
0answers
17 views
Can I set batch size to 1
I am trying to train a T5 (t5_large) transformer model on some data. Since it's out of cuda memory, I was forced to set ...
1
vote
1answer
32 views
Is it possible to train two neural networks combined with same output?
Say I have two kinds of input that I want my neural network to learn a possible from:
one 2D image
some 1D metadata about the image
This case and similar cases seem problematic to me because the two ...
0
votes
0answers
2 views
How are t+n predictions generated?
I'm trying to understand how ARIMA and other models generate or produce a prediction sequence.
I understand how they fit the model with each point with the training set, and then testing with the ...
1
vote
0answers
18 views
If I know specific pair of characters that model confuses in OCR task how can I fixe it?
I train OCR model to recognize cyrillic handwritten text. I know, for example, that it confuses very often 'Б' with '6'. How can I use this information to fine tune the model ?
Just in case, my ...
1
vote
0answers
78 views
Machine Learning Classification of Changes in a Time Series
I would like to classify the changes in the time series depicted in the figure below—representing an angular position history of a robot about the vertical axis—into increment/decrement of roughly (...
1
vote
0answers
14 views
Percentage of positive target values in dataset and the train/test split
Suppose that in my dataset of 100 observations, only 25 have a target variable equal to 1, while the other 75 have target variables equal to 0. Should the portion of target values that are positive ...
0
votes
0answers
17 views
Reporting cross-validated score and test score
Is this the right way to do cross-validation and testing of a machine learning model:
Split my data into train (85%) and test (15%)
Do hyperparameter tuning based on the train set
Do 10-fold cross-...
1
vote
1answer
14 views
Shall I present predictions on the (oversampled) training set as well?
I am dealing with an imbalanced classification problem and used oversampling on my training set to to predict on my testing set.
My PI insists on presenting evaluation metrics of the different trained ...
2
votes
0answers
45 views
Interview question: train/test error and "best" model
I recently had a puzzling interview question and I am wondering whether anybody can tell me the intended answer.
The question shows train and test error for three models plotted against the number of ...
0
votes
0answers
23 views
Two datasets have different imbalanced class. How to split?
I'm new in machine learning. Actually I have two datasets files as my scrapping results from different news webpages. I want to preprocess to just selecting relevant news. So I would like to perform ...
1
vote
1answer
524 views
How LightGBM deal with a new categorical value in the test set
Suppose I have the training data set $(X, y)$ where $X$ is my feature space $(x_1, \dots, x_n)$. Let $x_1$ be a categorical feature column. In my test set, if the feature $x_1$ takes a new categorical ...
0
votes
0answers
114 views
Train and test score - overfitting?
I have hourly time series data with a range of two years. I want to test my model when predicting my target variable (continuous) for a specific week. I'm doing the following:
Splitting my data into ...
6
votes
4answers
645 views
Train and Validation vs. Train, Test, and Validation
I am embarking on a new job that will give me the opportunity to do some cool machine learning stuff. I haven't touched this stuff on a deeper level since graduate school and I wanted to get some ...
0
votes
0answers
9 views
How to evolve an already trained model with new data
How to evolve a model when new data come out?
Suppose I have a model $M_1$ trained on data $D$.
After the model $M_1$ is trained, new data $T$ is available.
How to train a new model $M_2$ from $M_1$ ...
0
votes
1answer
269 views
Random Forest Regression and Overfitting
I did gridsearch with corss-validation on a trainingset to search for best hyperparameters for a Random Forest Regressor. And indeed the best parameterset gives good results in cross-validation (R^2 ~ ...
0
votes
0answers
50 views
How to compute TBATS forecast accuracy without specific test set?
I have used the TBATS model on my data and when I apply the forecast() function, it automatically forecasts two years in the future. I haven't specified any training set or testing set, so how do I ...
0
votes
1answer
34 views
Method for selecting training data and identify predictive genes in a survival model
I am performing a coxph survival analysis and want to split my data into a training and test set. During the training I would like to identify the "best" ...
0
votes
1answer
110 views
How is stratified sampling better than sampling equally from all classes while crossvalidating?
I can see that stratified sampling helps in maintaining the same class distribution in the training set as in the original dataset. However, my understanding is that ideally, the model should be ...
1
vote
2answers
43 views
Training a Neural Network on Music
I'm very picky about the music I like. I can't say I like a genre in particular because my tastes in music boil down to how a specific song sounds. For a while now, I've had a vision of creating a ...
0
votes
0answers
47 views
High precision, but low recall on training data. What could be wrong?
I am training RNN for timeseries binary classification. I observed that network has high precision, but low recall on both training and testing data. I tried multiple architectures, but same problem ...
