Questions tagged [machine-learning]

Machine learning algorithms build a model of the training data. The term "machine learning" is vaguely defined; it includes what is also called statistical learning, reinforcement learning, unsupervised learning, etc. ALWAYS ADD A MORE SPECIFIC TAG.

Filter by
Sorted by
Tagged with
0
votes
0answers
10 views

An efficient way to encode & embed tabular data of a video into a transformer?

So a little bit of a background: I have a folder which contains video files of lets say humans doing a certain action (i.e. walking) where each .2 seconds is documented in a ...
1
vote
1answer
40 views

Mathematically finding a threshold for deciding on rare word

I am implementing spell-correction facility as a preprocessing step of for a text classification project. For this reason I have to make a knowledge-base where I shall be putting all the words along ...
-1
votes
0answers
20 views

Variational Inference: compute the an expectation of log likelihood

I want to compute the $\mathbb{E}[\text{log} \ p(\text{data}| z)]$, where $z$ describes the probability of a coin coming up heads in trials. According to this post, ...
0
votes
0answers
15 views

Why Difference in ROC-AUC Scores?

I have computed ROC-AUC Scores via both true labels and predicted labels as well as via true labels and predicted probabilities.But there is difference between scores when computed separately via ...
1
vote
2answers
65 views

Can I compare the output probabilities of two machine learning models?

I'm sorry if this is a silly question. Suppose there are two logistic regression models $M_1$ and $M_2$ trained on the same (or similar) dataset, and their outputs of given input $x$ are $P_{M_1}(y \...
0
votes
0answers
9 views

Unsupervised anomaly detection and classification with event (log) data

I am trying to detect anomalies in a large set of user log events, where most users would be considered “good” and a small minority would be considered “bad.” There are hundreds of event types, which ...
1
vote
0answers
13 views

Online vs Offline Triplet Selection in FaceNet

I have been reading FaceNet. In the Triplet Selection section, it is written Generate triplets offline every n steps, using the most recent network checkpoint and computing the argmin and argmax on a ...
0
votes
0answers
3 views

Information neutral feature transformation

In a supervised learning task, with dataset $X, y$, is there any non-trivial feature transformation $t(X) \to X'$ that is guaranteed to be neutral? By neutral, I'm referring to, with the same model, ...
0
votes
0answers
18 views

Longitudinal Clustering

I wondered if anyone was aware of research and corresponding R packages based upon unsupervised clustering in two different states. For example, suppose I have a panel data sample with 12 ordinal ...
0
votes
0answers
5 views

Reporting multiclass macro R/P/F1 when classes are absent from real world data

I have developed a hierarchical classification framework for my current project using a 'local classifier per parent node approach' (as described in https://link.springer.com/article/10.1007/s10618-...
0
votes
0answers
24 views

Variational coin tossing from scratch: calculation of the expected log likelihood

I'm working my way through this tutorial about variational inference for a coin tossing. Let's say the probability of the event head is denoted by ...
-2
votes
1answer
20 views

Custom preprocessing using piplines [closed]

I have searched a lot for this issue but unfortunately came up with nothing. Usually in a ML model, during preprocessing, we use Pipelines and ...
1
vote
1answer
68 views

How to prove Mean Squarred Error (MSE)

I would like to prove this equation of Mean Squared Error (MSE): m is the number of training instances. X is a m × n matrix containing all the feature values (excluding labels) of all instances in ...
1
vote
1answer
20 views

Interpreting NaN values for precision in Confusion Matrix

Please refer to the confusion matrix here: https://imgur.com/a/Iq1epre Would I get precision values of NaN because of 0/0 in the right most columns? Is that even possible? How should I interpret this? ...
0
votes
0answers
17 views

Running Variance Inflation Factor(VIF) iteratively for a large number of features taking a long time

I am trying to do logistic regression on a dataset with 1,500 features that are very multicolinear. I care about interpretability of coefficients so I am running a VIF calculation on all columns and ...
0
votes
1answer
27 views

Finding out the generative log-likelihood $\log(p(x|t))$

The question ask to build classifiers to label images of handwritten digits. Each image is $8$ by $8$ pixels and is represented as a vector of dimension $64$ by listing all the pixel values in raster ...
1
vote
1answer
10 views

Two basic questions about icp (iterative closest point) algorithm

I am trying to learn shape analysis and a part is learning icp. I have many confusions but for now I have two basic questions: Does the point clouds need to have the same number of points for icp? ...
2
votes
0answers
58 views

Validation loss falls but train loss remains constant? [closed]

My validation loss (left) falls to near 0, while my training loss (right) remains basically unchanged (gradient step is on the abscissa). This is the opposite of the typical error in which train loss ...
0
votes
0answers
65 views

"Automatically" detect biased subsets in probability distribution

Background: Suppose we have a model generating probabilities conditional on a state vector - for simplicity we can just assume the outcome is 0 or 1 (imagine for example simple logistic regression): $...
0
votes
0answers
21 views

systematically determine which model is better on a specific dataset

In general, there is no silver bullet to choose a better machine learning model for all datasets. However, I'm wondering, what if we fix a dataset including all the train/test set ahead of time. Given ...
0
votes
0answers
43 views

What is so novel about Machine learning in general and nueral network in particular? [duplicate]

To give a brief one-line description of machine learning: It is basically a function approximation given sample and hypothesis class. But this question is already tackled by statisticians (parameter ...
1
vote
0answers
17 views

Preprocessing for the final model to be deployed [duplicate]

Typically for a ML workflow, we import the data (X and y), split the X and ...
2
votes
1answer
36 views

Should I use every Dicom slice of a Dicom series?

I have a skull fracture CT scan dataset, consisting of fracture or normal cases. My question is: Let's say patient-1 has a skull fracture, and his CT scan has 300 Dicom slices. Now should I label ...
2
votes
1answer
10 views

Kernel pca and kernel svm

PCA is a way to reduces dimension and complexities, but is it ok to use kernel PCA with radial basis function and then use kernel SVM using the same.
1
vote
1answer
18 views

Would decision tree propagate error?

Given a regression dataset $X,Y$. Suppose there are two different decision tree (CART) $T_1, T_2$ fitted from it. Each using different feature encoding method. And we get two different tree. And there ...
1
vote
0answers
20 views

How to choose the balance between model fit vs AUC (diagnostic accuracy)?

I would like to know how we can choose between model fit (calibration) vs AUC when building the predictive model. For example, if I have one predictor which improves the model fit but results in a ...
1
vote
0answers
34 views

Why do linear bandits use ridge regression to estimate parameters?

I’m implementing an adaptive experimental design where arms are assigned according to the posterior probability that they are the best arm. I’ve noticed in several articles that people use ridge ...
1
vote
0answers
19 views

How can one feed all of the input to an RNN, and then get all of the output from it?

When reading papers, a common concept is delaying the output of RNNs to after seeing all of the input. E.g., the neural Turing machine paper uses this technique, together with a simple identity ...
0
votes
1answer
16 views

Decision Trees and SHAP Values

I've recently been using some (optimal) decision trees methods in R, such as 'evtree' and 'iai.' Both of these provide really nice interpretable plots. And out of the 12 covariates I have in my model, ...
0
votes
0answers
14 views

Minimizing the Variance in Compressed SGD

I have been reading this paper [1] and [2]. There is a statement that reads as follows: The variance between compressed gradient and uncompressed gradient must be made as small as possible because ...
1
vote
0answers
22 views

Federated Learning and Continual Learning: Non-IID Learning

I have some exposure to federated learning and continual learning which are non-iid learning instances [1] and [2] I was wondering can we state the following: Federated learning is when the dataset ...
5
votes
2answers
334 views

matrix-calculus - Understanding numerator/denominator layouts

Consider the following machine-learning model: Here, $J = \frac{1}{m} \sum_{i = 1}^{m} L(\hat{y}^{(i)}, y^{(i)})$, and $m$ is the number of training-examples. While performing reverse-mode ...
2
votes
0answers
32 views

Help interpreting image prediction

Could someone tell me why my predicted result ("Predicted Heatmap") has "ghost layers" and gray background? What can I do to improve my model? **What I've done to the images ** ...
0
votes
0answers
17 views

Pearson Residual Goodness-of-fit test (Updated)

I am running a logistic regression predictive model with death (sta) as the binary outcome variable, and age (continuous variable), and cancer status (variable can; categorical variable) as predictors....
0
votes
1answer
44 views

Can I calibrate to 100% of my sample in ML regression?

I have a standard ML regression model trained on 80% of my data with 20% saved for testing. I want my model to match my full sample as best possible. Can I multiply my outputs by mean(observations ...
2
votes
2answers
322 views

Need help with understanding Decision Trees [closed]

I am struggling to understand how decision trees work. I understand that you need to calculate the Gini coefficients for the sample features and that's how leaves are chosen. My issue is that I don't ...
1
vote
0answers
32 views

Feed-Forward ReLU Networks as a matrix multiplication

When reading papers, Feed-Forward NN are often formalized as follows: $$\Phi(x) := \sigma(W_L\cdots \sigma(W_2\cdot \sigma(W_1x))\cdots) $$ i.e., the ReLU activation function $\sigma$ applied to the ...
0
votes
0answers
13 views

How to improve accuracy of natural's Naive Bayes classifier?

Note: I'm new to Machine Learning and NLP. This is my first project in this field. I'm using NaturalNode/natural (https://github.com/NaturalNode/natural) to build a chat bot to help my users with ...
0
votes
0answers
4 views

Masked language models: can you train on remasked data?

Masked language models like BERT and friends are trained on the task of predicting words removed from input text. Normally, this text is removed at random from some training data. As far as I can tell ...
0
votes
0answers
19 views

How much taking uncorrelated features for a Machine Learning problem is important?

I imagine that taking correlated features for some kind of ML algorithms is no useful (as for a Linear Regression) but it doesn't not hold for all algorithm in general. There are algorithms that don't ...
0
votes
0answers
20 views

One class SVM and centered data

I understand that the one class SVM try to separate the normal training data point from the origin. My guess is that, if we centered the data in a normalisation step, the OCSMV will works poorly since ...
1
vote
1answer
74 views

Cross validation and hyperparameter tuning workflow

After reading a lot of articles on cross validation, I am now confused. I know that cross validation is used to get an estimate of model performance and is used to select the best algorithm out of ...
1
vote
1answer
35 views

Evaluating Probabilistic/Bayesian Forecasts - PIT Values & How to Generate

Suppose you are modelling a linear regression $y_i = \alpha + \beta x_i + \epsilon_i$, in probabilistic terms: $$ \mu_i = \alpha + \beta x_i, $$ $$ y_i \sim \mathcal{N}(\mu_i, \sigma). $$ For each ...
0
votes
0answers
7 views

How to interpret the variable importance varImp() when training a LASSO/Ridge regression using the library caret and method = "glmnet"?

I have trained an elastic net regularized model and left with my top two variables - both factors. • How can I interpret the importance of each one? • Should I train a new linear model including only ...
0
votes
0answers
24 views

Explainable AI - Traditional ML algos

In my work, I mostly use traditional algorithms such as Logistic regression, Linear regression, SVM, Naive Bayes, Random Forests, Decision Trees and Boosting etc to analyze data and make predictions. ...
0
votes
0answers
9 views

When do we need to log transform predictors for logistic regression model? [duplicate]

I am interested in knowing when we should log-transform the predictors for the logistic regression model? My predictors are highly skewed but I read about some materials online that we don't need to ...
1
vote
0answers
15 views

Finding causal Inference from sentiment analysis

I am conducting a sentiment analysis on thousands of social media posts by unemployed manufacturing workers to see how online sentiment of the group members I am analyzing has changed after an ...
2
votes
1answer
74 views

How many predictors can I include in my logistic regression model

If I am dealing with a small sample size (n = 48; n = 29 have disease vs n = 19 without disease), what are the maximum numbers of the predictors I can include in my multivariable logistic regression ...
0
votes
0answers
7 views

How to determine what predicted probability to use in the Risk stratification table?

I am new to the prediction model and would be very grateful for the advice regarding the likelihood ratio test. I want to construct a risk stratification table like the one in this study (see Table 4)....
0
votes
0answers
10 views

Study Resources for optimizing(increasing/decreasing) a certain feature using AI

I am looking for resources that talk about how we can optimize(increase/decrease) a certain independent variable using other independent variables and the dependent variable. For example, we have 3 ...