Questions tagged [training]
Training is the part of machine learning whereby a model is "trained" on a define portion of a dataset to learn attributes and statistical features of the data. It's counterparts are called Testing and Validation. After training a model is tested and validated on another portion of the dataset.
680
questions
0
votes
1
answer
32
views
RNN for continuous, real-time learning without pre-training
I am learning ML and I'm trying to solve this problem
Create a rock paper scissors game where the AI is able to beat the player more than 50% of the time.
My initial intuition was to use an RNN with ...
0
votes
1
answer
28
views
Where should I stop training if I want to bag models
Let's say I have a clear case of overfitting where my loss curves look like this (x axis are iterations):
Now I would like to try bagging to reduce the variance, where should I stop models training?
...
- 111
0
votes
1
answer
30
views
How to train a regressor model on data that has duplicate subjects but different records for each?
I am working on a dataset that is as follows (just an example):
prop_subj
prop_comp
bed_subj
bath_comp
sqft_subj
sqft_comp
A
B
2
1
1002
1006
A
C
2
2
1002
1075
A
D
2
2
1002
1000
B
G
2
1
1002
978
...
1
vote
1
answer
59
views
Is backpropagation applied every layer the same?
For example, I have layers that are pretrained. But while predicted, the loss is very high. But not because of pre-trained layers. Because of not pretrained layers. Will every layer be affected by ...
- 91
1
vote
1
answer
97
views
Avoid leakage in NLP extraction
What is best practice for applying traditional NLP extraction techniques a pre-processing for ML models?
Given a pipeline:
Collect raw data.
Parse full data set with a variety of traditional NLP ...
- 435
0
votes
0
answers
33
views
Different results between training and evaluation phase on the same data
I have trained a CNN and in the training phase I obtained an accuracy of 36.5%.
If I call model.predict() on the same test data of the training phase I only obtain ...
0
votes
1
answer
175
views
Weird consequence of not freezing layers in Neural Network
I was researching about "why are we freezing layers" and I came across the answer says "to not lose the information of pre-trained model" But; we are just freezing early layers (I ...
- 91
0
votes
0
answers
245
views
How ca I reshape y_train , y_validation from train_generator?
I retrained ResNet-50 for iris flower classification in tensorflow using the following code:
...
- 125
0
votes
1
answer
25
views
How to proceed when training data change frequently (in production)?
I'm working with a Recommendation System that would take as parameters a bunch of "tweets" a user see during his navegation on a mobile app. Every tweet has a property, like a category (...
0
votes
1
answer
19
views
Distribution / number of examples
Imagine you want to predict if a picture is showing a cat or not. First you train your ML algorithm with examples of pictures of cats and dogs and it works. But then you want to train it to also ...
0
votes
2
answers
716
views
Why is my DCGAN not converging?
I'm training a tf DCGAN on the MVTec hazelnut dataset and I found some difficulties. The problem is that after a lot of epochs the generate does not produce some quality images.
My model is the ...
- 1
0
votes
2
answers
2k
views
0
votes
0
answers
16
views
Training with code scripts to achieve a specified goal?
I am quite new to machine learning and therefore need to ask if some ideas might be possible.
Imagine an application that is managing the state of an Actor by ...
- 101
0
votes
0
answers
34
views
What does the learning curve indicate?
I am training a deep learning model for traffic prediction. When I use 10 months of data for training (validation split: 10%) and 2 months for testing. The loss curve looks like this:
.
and the ...
2
votes
0
answers
303
views
Your input ran out of data; interrupting training
I am trying to train ResNET50 for dogs and cats classification (Tensorflow2.3) using the following code:
...
- 125
1
vote
1
answer
102
views
Getting equal distributions of data from different input sets
I am new to ML. I am trying to create a training dataset that is equally distributed between multiple lists, each of which have a different kind of data. How can I do this? I looked into ...
- 11
1
vote
0
answers
75
views
Partial fit with tslearn clustering methods
Is there a way to use partial_fit with tslearn clustering methods like TimeSeriesKMeans? I ...
- 11
0
votes
0
answers
84
views
training loss decreases but training accuracy is also decreasing with epochs
I am working on the classification problem where by I am having a hinge loss function + other loss terms to optimize for which the input is the output from tanh layer at the end. But I can't reveal ...
- 121
0
votes
1
answer
266
views
Training deep neural networks with ReLU output layer for verification
Most algorithms for verification of deep neural network require ReLU activation functions in each layer (e.g. Reluplex).
I have a binary classification task with classes 0 and 1. The main problem I ...
- 1
2
votes
1
answer
49
views
Is batch size of 1 a valid choice for a very deep neural network with high memory requirement?
I am training a very deep neural network (Panoptic-DeepLab) with a ResNet34 backbone on Google Colab on CityScapes dataset for Panoptic Segmentation, and noticed that, with a big crop size, the batch ...
- 121
1
vote
0
answers
99
views
Relation between batch size, number of steps, and learning rate
Taking alphazero training setup as a reference:
700k total steps
batch-size of 4096
initial LR of 0.2
What would be an equivalent setup for a batch-size of 1024? ...
- 111
2
votes
1
answer
725
views
Why we do random sampling when we select the training set?
The usual workflow when building a machine learning model starts with random splitting the data set into training and test set.
What I can't understand is why we do this. For example lets say we have ...
- 181
1
vote
2
answers
41
views
Scalar predictor - is it better to have a lot of training data that is less precise? Or fewer training data that is more precise?
I am quite new to this neural network stuff, so please bear with me :)
TL;DR:
I want to train a neural network to predict a scalar score from 32 binary features. Training data is expensive to come by, ...
- 13
0
votes
1
answer
1k
views
KL divergence loss first decreases and then increases in VAE training
I am training a VAE on CelebA HQ (resized to 256x256). The training is going well, the reconstruction loss is decreasing and reconstructions are also meaningful. But, the problem is with KL divergence ...
1
vote
0
answers
14
views
State-of-the-art techniques for regularizing Neural Networks?
For regularizing neural networks, I'm familiar with drop-out and l2/l1 regularization, which were the biggest players in the late 2010's.
Have any significant/strong competitors risen up since then?
- 111
1
vote
1
answer
29
views
Can depth be used as a feature when predicting rock type from well log data?
I am trying to predict the lithofacies, i.e. the rock type, from well log data, a project very similar to the one described in this tutorial.
A well log can be seen as a 1D curve tracking how a given ...
- 195
2
votes
1
answer
84
views
Is it recommended to train a NER model using a dataset that has all tokens annotated?
I'd like to train a model to predict the constant and variable parts in log messages.
For example, considering the log message: Example log 1, the trained model ...
0
votes
1
answer
664
views
Training loss decreasing while Validation loss is not decreasing
I am wondering why validation loss of this regression problem is not decreasing while I have implemented several methods such as making the model simpler, adding early stopping, various learning rates,...
3
votes
1
answer
932
views
How are parameters selected in cross-validation?
Suppose I'm training a linear regression model using k-fold cross-validation. I'm training K times each time with a different training and test data set. So each time I train, I get different ...
- 233
0
votes
2
answers
82
views
NNs for fitting highly oscillatory functions
in a scientific computing application of neural networks, I have to maximize several neural networks with scalar output with respect to a target/loss function (coming from a weak form of a PDE).
It is ...
- 1
0
votes
0
answers
41
views
What will happen if we train a model on a dataset sorted by class
Suppose we have a dataset of two classes (0 and 1) divided into over 12k mini-batches where the first half of the dataset (over 6k mini-batches) belong to class 0, and the other half belongs to class ...
0
votes
1
answer
45
views
Coefficients values in filter in Convolutional Neural Networks
I'm starting to learn how convolutional neural networks work, and I have a question regarding the filters. Are these chosen manually or are they generated by the network in training? If it's the ...
0
votes
2
answers
95
views
Standardization in combination with scaling
Would it be ok to standardize all the features that exhibit normal distribution (with StandardScaler) and then re-scale all the features in the range 0-1 (with <...
- 119
1
vote
1
answer
386
views
Training data for anomaly detection using LSTM Autoencoder
I am building an time-series anomaly detection engine using LSTM autoencoder. I read this article where the author suggests to train the model on clean data only in response to a comment. However, in ...
1
vote
2
answers
30
views
Cross Validation after using train-test to decide optimal algorithm to use?
I am interested in training different algorithms on a data set and observing performance metrics.
Currently, my approach is to train different algorithms on train data, and then evaluate performance ...
- 11
1
vote
2
answers
2k
views
How to train-test split a timeseries?
I have a dataset consisting of multiple timeseries for multiple users. So per user I have multiple timesteps, a value to predict per timestep and a list of features per timestep. I am currently ...
- 131
0
votes
1
answer
283
views
What causes explosion in MSE when training?
I (probably) well overfitted/overtrained a model. But I was just curious as to what might cause this type of behaviour. I carried on training (Epoch 1/50 is not the first epoch of training this model)....
- 91
0
votes
1
answer
29
views
How is model evaluation and re-training done after deployment without ground truth labels?
Suppose I deployed a model by manual labeling the ground truth labels with my training data, as the use case is such that there's no way to get the ground truth labels without humans. Once the model ...
- 133
0
votes
1
answer
129
views
Training Loss or Validation Loss for Hyperparameter Optimisation
When performing HO, should I be looking to train each model (each with different hyperparameter values, e.g. with RandomSearch picking those values) on the training data, and then the best one is ...
- 91
0
votes
1
answer
2k
views
TypeError: bad argument type for built-in operation
I have one dataset containing images X of type ( numpy array) and one target csv file as Y which has counts of cells (type : pandas dataframe, that I have converted to numpy array), both are now read ...
- 5
2
votes
2
answers
100
views
Does it makes sense to train the model on whole data?
Suppose I am training an lstm model on a stock price data.
So for first iteration say I have trained it on 80% of data and then tested it on rest of the 20% data and got the rmse value.
Now after this ...
- 157
1
vote
1
answer
1k
views
1
vote
0
answers
15
views
How to add training costs to the variables of a model? [closed]
I am dealing with the Breast Cancer dataset and I want to include costs to the variables, trying to minimize the training costs and maximizing the accuracy. The costs for each variable are as follows
<...
- 304
0
votes
1
answer
39
views
Machine learning problem - train and test with different classes
I have a datasets composed like this:
library(caret)
data <- iris
train <- data [1:75,]
test <- data[76:150,]
So, I have 3 classes in total but:
Train: ...
- 481
1
vote
2
answers
62
views
How to generate test set with no data-leakage using multiple columns
I am developing a fraud detection algorithm. Among other things, my dataset contains the phone number, email address and a few other fields that should uniquely identify a user (let's call them "...
- 181
0
votes
2
answers
884
views
Why label encoding before split is data leakage?
I want to ask why Label Encoding before train test split is considered data leakage?
From my point of view, it is not. Because, for example, you encode "good" to 2, "neutral" to 1 ...
- 73
0
votes
1
answer
276
views
What the difference between a flattening validation curve and one that increases again?
I know that we monitor the validation loss to investigate overfitting. I am familiar with the validation curve that first decreases and then increases again. The increasing part means the model starts ...
- 113
0
votes
1
answer
134
views
how can i set learning rate for big data?
will it need more epochs for training or it is not a necessary and what is the learning rate I should set for this data with optimizer adam?
- 25
1
vote
0
answers
18
views
Should the model be defined again before training it to new data?
I wanted to fit the LSTM model on new data set in a loop so I have implemented it like this
...
- 157
1
vote
0
answers
49
views
How to include the sudden peaks/bursts in LSTM based time-series model's training
I am using LSTM for time-series prediction whereby I am taking past 50 values as my input. Now, the thing is that it is predicting just OKish, and not doing the exact prediction, especially for the ...
- 131