Skip to main content

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
4 votes
1 answer
2k views

In Machine Learning, how does getting more training examples fix high variance $(Var(\hat f(x_{0})))$?

I don't believe that (Why does increasing the sample size lower the variance?) appropriately handles my question! The linked questions explains why any addition of random variables (all iid) produces ...
0 votes
0 answers
14 views

Customized distribution fitting

I have a sample of continuous surgery durations and want to fit a distribution to this sample to generate random numbers accordingly. Using the scipy library in ...
0 votes
0 answers
15 views

Threshold selection in chain classifier

Imagine a set of $m$ classifiers and $N$ precedents, each assigned a threshold $\theta_{i,j}$, meaning, if threshold $t_j$ on classifier $j$ for precedent $i$ is bigger than $\theta_{i,j}$, the ...
1 vote
0 answers
47 views

Does artificially balancing outcomes in regression lead to poor calibration? If so, how to show the poor calibration?

In "classification" problems, it is common for there to be unbalanced-classes. To combat what appears to be a problem (though I would argue that it usually is not a problem), it is common ...
1 vote
0 answers
4 views

NER With Custom Tags, How to Approach

I am building a "field tagger" for documents. Basically, a document, in my case something like a proposal or sales quote, would have a bunch of entities scattered throughout it, and we want ...
1 vote
0 answers
18 views

I have a dataset with 18 biomarker features and a target variable. I want to find the features which are having the biggest impact on the target [closed]

I Have some disease biomarker datasets that contain 18 biomarker readings from different samples and a target variable which shows presence or absence of disease (features are both categorical and ...
1 vote
2 answers
51 views

Is duplicating dataset an augmentation?

For a very small dataset, there is a lot of overfit in the random forest regressor model. I have removed extraneous data, scaling and feature selection, but overfit is still there. The oversampling ...
0 votes
1 answer
575 views

quantify ML algorithm accuracy, reducing false negatives (i.e. breast cancer detection)

I'm doing some anomaly detection, basically classifying stuff as normal (0) or aberrant (1) As with all anomalies, they are rare, so during the train/test phase it's not good enough for me to just ...
0 votes
1 answer
23 views

Should I Use Regularization in Univariate Logistic Regression for Diagnostic Methods Comparison?

I am comparing two diagnostic methods, Method 1 and Method 2, where Method 2 is considered the gold standard. I am using Method 1 to predict the Method 2 using logistic regression. My dataset contains ...
2 votes
1 answer
798 views

Model Stability Metrics and Solutions

I encountered the term of "stability" in learning theory, well described on this wiki page and its references. The idea described in this page deals with the notion of how sensitive the ...
4 votes
1 answer
919 views

Likelihood of Linear Discriminant Analysis compared to logistic regression

I've come across an interesting exercise. We are given four classification models for binary response and a $d$-dimensional independent variable: A Linear Discriminant Analysis model where the ...
-1 votes
0 answers
48 views

What are some SOTA algorithms for single hyperparameter optimization while training Deep Neural Networks?

I would like to optimize a single hyper-parameter while training a deep neural network. Let's say it is the learning rate of the network. What algorithms should I use to optimize the process? A ...
0 votes
2 answers
104 views

How to generate 95% prediction interval around predictions from ML model?

I have predictions from an ML model and would like to generate 95% prediction intervals around each prediction generated from the model such that I can claim that these are the plausible range of ...
1 vote
3 answers
55 views

Can I apply data augmentation to the test set?

I'm working with a dataset of 102 rows (tabular data), from which I'm using 91 for training and 11 for testing. I'm using data augmentantion through the addition of gaussian noise for the training set....
0 votes
1 answer
243 views

Accuracy of SVM prediction

I'm trying to build a text classification model with SVM. The training data set consists of 100 string records with a one-to-one mapped response variable which is also a string. I can't split the data ...
1 vote
0 answers
17 views

Why is the threshold term incorporated into the weight vector in linear classifiers?

In the context of linear classifiers, such as the perceptron or logistic regression, I understand that the decision boundary is defined by a linear combination of input features and weights, plus a ...
3 votes
1 answer
608 views

Difference between cumulative gains chart and cumulative accuracy profile for binary classifier

I am confused about the following: Here I find the definition of cumulative gains chart as the plot of x-axis: the rate of predicted positive y-axis: true positive rate It turns out that we can e....
2 votes
2 answers
41 views

Defining clinical follow-up: Fixed Period vs. Maximum Duration

We are retrospectively analyzing data of around 1100 patients operated between 2017 and 2023. We analyzed follow-up documentation until 2024. This means that patients operated at a later date will ...
0 votes
0 answers
36 views

Environmental filtering versus spatial resampling in species distribution modeling

I am building species distribution models using machine learning models based on GBIF data (presence-only data) and working on a very large spatial scale, encompassing all of North America. Before ...
1 vote
1 answer
313 views

Least squares optimization with expensive model and many parameters

I have a physical model which takes $\sim50$ parameters and gives $\sim 2000$ outputs taking tens of minutes to run. I need to optimise these parameters to give outputs as close as possible to data by ...
1 vote
1 answer
387 views

What is the best way of creating new features in a dataset?

I recently started working with sklearn, and found myself creating new features often (new features with K Bins, with various Encoders etc.). What I noticed though, is that is very difficult to ...
4 votes
2 answers
277 views

What is the Gold Standard for Evaluating the Posterior of a Bayesian Regression Model?

Let me explain my meaning & the context: I mean evaluating the correctness of the posterior (e.g. for approximate Bayesian inference methods). I care mostly about Bayesian deep learning, I'd like ...
4 votes
1 answer
199 views

Link between Cross-entropy and MLE

There are numerous material that show the relationship between MLE and cross-entropy. Typically, these are the steps taken to show the relationship for a I.I.D data generating process $D = (X,Y)$: $$ ...
17 votes
2 answers
8k views

How is the Upper Confidence Bound derived?

I came across the formula for obtaining the upper confidence bounds on the k-armed bandit problem: $$c\sqrt{\frac{\text{ln} N_i}{n_i}}$$ where $n_i$ is the number of samples we have for this ...
0 votes
0 answers
16 views

Independance in Bias Variance decomposition [closed]

In the second line of derivation we use independance of $\varepsilon$ and $\hat{f}(x)$, but which hypothesis lead to this independance result ? Is it because all observations are iid ?
1 vote
2 answers
831 views

Multi Level / Hierarchical Time Series Models in Python

I have data with different leaders and their performance score over a period of time. I need to analyse this time series data where data for each leader should be considered different time series data....
1 vote
1 answer
222 views

Multi stage neural net

I'm going to use my specific use case here, but I'm curious about this AI structure in general. I want to create an AI to play Rocket League (a video game). For context, in Rocket League you control ...
0 votes
1 answer
219 views

Overfitting the (non-nested) cross validation set

This is a follow up query to this query In one line: I wish to understand why is it that we will severely over fit the cross validation set (and hence need nested cross validation to correctly account ...
0 votes
0 answers
18 views

Why does the square term get omitted in Gradient derivation of parameter θ-th

I get it that my question may sound a bit sophisticated or overwhelming, but it's pretty straightforward when you read the image below. As you can see, the square ^2 completely dissipates, despite ...
6 votes
2 answers
2k views

How do I find multiple change points in an online dataset?

I am trying to develop a Python based script connected to a SQLite3 database to identify distinct system changepoints in an "online" data stream. A change point must be identified in less ...
2 votes
1 answer
2k views

Sum of Squared Error Chi-Square distribution degree of freedom in Multilinear Regression

In this link it says that $Y$ variables has zero covariance (because covariance matrix has only diagonal terms) which implies they are independent. Actually in linear regression $Y$ takes its ...
4 votes
2 answers
7k views

About cross-validation for machine learning

Assume I have 1000 samples of data. I split the data randomly into training and test sets of size 800 and 200, respectively. Now, I train a classifier using the training set, and then evaluate the ...
-1 votes
0 answers
11 views

Is Analyzing Post-Fire Vegetation Recovery with NDVI Values Considered an Interrupted Time Series Analysis? [closed]

I'm working on a project that involves analyzing the recovery of vegetation after a wildfire using NDVI (Normalized Difference Vegetation Index) values. My goal is to assess how vegetation regenerates ...
0 votes
1 answer
259 views

How to Predict the sales of all the items, offered in all the countries

I am working on a task to predict the sales of all the items offered in all the countries. The sales are aggregated on a daily and country level. Each Item has a history of past sales and prices for a ...
3 votes
1 answer
409 views

Is it possible for a reinforcement learning agent to create or generate additional features

Based on what I've read, the best model-free reinforcement learning algorithm to this date is Q-Learning, where each state, action pair in the agent's world is given a Q-value, and at each state the ...
2 votes
1 answer
357 views

Regularization versus feature reduction in species distribution modeling using Maxent

I am wondering if there is a need to set the beta multiplier in Maxent (species distriubition modeling approach) if one is also reducing features using a contribution threshold. I have seen a number ...
0 votes
2 answers
226 views

Should I separate my data into different batches and then perform tsne on each batch?

I have a very huge dataset and required to reduce the embedding of 768 dimension to 128dimension with TSNE. Since I have more than 1million rows, it takes more than weeks to complete dimension ...
0 votes
0 answers
27 views

A faster way to choose the right number of features and the right parameters [closed]

For a very small dataset with less than a hundred samples, gridsearch does not give us the desired results and has overfit. But when I perform hyperparameter tuning of the model manually and also ...
0 votes
0 answers
21 views

How Random Forest handle missing value in sk-learn? [duplicate]

What is the technic used in Random Forest Regressor from scikit-learn to handle missing value ? First I thought that a Random Forest regressor was able to natively handle missing value during training ...
2 votes
1 answer
338 views

SVR with combination of kernels

I'm trying to use Support Vector Regression (SVR) to model a trend. I have a dataset with dependent variables $Y$ and features $X$. Suppose for simplicity that both $Y$ and $X$ take real values from $-...
0 votes
0 answers
7 views

Using scale pos weight and non 0.5 cut off score for a look-alike model

I'm working on a classification problem where I'm trying to identify look-alikes of Class 1 in Class 0. Class 1 and Class 0 are established based on type of product customers use. Basically, Class 1 ...
4 votes
2 answers
961 views

Expected value of error in neural network

I wanted to take a look at the properties of the error vector that is propagating during backpropagation. The error vector $\boldsymbol{\delta}$ at layer $i$ is nothing more than the derivative of ...
0 votes
0 answers
26 views

Test/validation set

I've been having a discussion with colleagues and wanted to seek your input. If I'm using holdout and cross-validation to build and test my models. In this process, the training set is used to tune ...
2 votes
0 answers
23 views

Using whole training set for choosing model

I am working on a classification problem with what I understand as a big dataset. I have first of all splitted it in my "train" dataset and the "test" one. (Actually I am convinced ...
31 votes
5 answers
86k views

What impact does increasing the training data have on the overall system accuracy?

Can someone summarize for me with possible examples, at what situations increasing the training data improves the overall system? When do we detect that adding more training data could possibly over-...
0 votes
1 answer
3k views

Feature Selection for Multivariate Multilabel Time-series Classification

This is my first post and I am a beginner. I am working on a dance recognition project, where I collected skeletal data from one dancer performing 5 different gestures. My goal is to detect any pre-...
1 vote
1 answer
380 views

What is the best k in kmeans clustering [duplicate]

I did clustering on a dataset of real-world patients and since the best way to choose the amount of clusters in KMeans clustering is Elbow method and the Silhouette method, I conducted those two and ...
1 vote
0 answers
42 views

Optimal performance measures for species distribution models with presence and pseudo-absence data

I am currently building species distribution models (SDMs) (or ecological niches) using machine learning algorithms to predict the potential spatial distributions of animal species based on ...
6 votes
1 answer
1k views

Ensemble classifiers trained using different sets of features

Background I have a dataset, each record in this set is represented by 2 different sets of features. Let's say feature set A and feature set B. I have trained classifiers using feature set A and ...

1
2 3 4 5
406