Questions tagged [feature-engineering]

Feature engineering is the process of using domain knowledge of the data to create features for machine learning models. This tag is meant for both theoretical and practical questions regarding feature engineering, excluding questions asking for code, that would be off-topic on CrossValidated.

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

Different features for different models

Is it considered 'fair' to use different input features when comparing forecasting accuracy between models? For example, adding feature x1, x2 and x3 into a random forest while adding x2, x3 and x4 ...
0
votes
0answers
12 views

How to combine all lexical features and then run on a model in scikit learn?

I have a news database with columns 'label' and 'text'. I'm trying to perform a supervised learning in predicting the label of a given text. I can extract word and char n-gram features separately ...
2
votes
1answer
50 views

Classification on Dataset with multiple rows per person

I have a data set where there are multiple rows per person. If person1 has 3 rows, out of 10 features, only couple of them change. The remaining features have the same values repeated. There are ...
1
vote
1answer
14 views

Engineering features that depend on more than one data point (classification with gradient boosting in particular)?

So I'm working on the Titanic data set (predicting survival of passengers), and would like to add a feature that indicates whether a given passenger's family survived or not (using the known training ...
1
vote
1answer
63 views

Can anybody understand this specification of a feature for a horses preference to a condition?

I have become interested statistical analysis of sports and came across a horse racing paper: "Computer Based Horse Race Handicapping and Wagering Systems: A Report" (found at: https://www....
0
votes
0answers
10 views

feature selection using RFECV vs trees?

the RFECV takes the estimator and finds which features are important according to that estimator. But, it gets slow when we are dealing with data with many features. But, if we use trees they give us ...
0
votes
0answers
28 views

How to implement feature selection in machine learning? [closed]

I am working with some files which have around 15-20 features where these features are statistical features and I was implementing the several Classification models to classify different cases. I ...
1
vote
1answer
22 views

feature hashing for high cardinality

when we are applying feature hashing in sklearn it asks us what should be the dimensions of feature required for us. If we decrease too much there will be more collisions which are not good. And we ...
1
vote
1answer
12 views

How to replace scalars with vectors in simple models, such as classification of sentences where 1-hot encoding is replaced with word vectors

I have a problem, which seems simple enough, but I don't know how it is solved in the industry. This has to do with the machinery of feeding data to a model, rather than trying to figure out the best ...
0
votes
0answers
5 views

Does changing reference variable of design matrix reduce multicollinearity?

It is stated here by Paul Allison that, if multicollinearity occurs within the design matrix of a categorical variable, multicollinearity can be reduced by setting the category with higher share of ...
0
votes
0answers
11 views

Month over month prediction for credit defaulters

I'm working on a month over month prediction using credit historical data. I've created more than 600 variables for this prediction including customer's delinquency in last 3/6/12/24 months etc. I've ...
0
votes
1answer
26 views

Scikit learn models gives weight to random variable? Should I remove features with less importance?

I do some feature selection by removing correlated variables and backwards elimination. However, after all that is done as a test I threw in a random variable, and then trained logistic regression, ...
1
vote
1answer
15 views

How to productionize a k-fold target-encoded feature?

I am attempting to build a model that has many predictors which are both categorical and large in cardinality. Target encoding looks to be a good solution for including these features, but I'm unsure ...
0
votes
0answers
12 views

What is the effect of Standardization on tree based models?

I have read that tree-based model such that Decision Tree, RsndomForests, Xgboost, etc do not require the standardization of data before feeding it to model but when I was working on working on ...
0
votes
0answers
3 views

Vector and curve in random forest feature

I am studying movement in images, and I need to classify them. In order to do so, I've created sequences, that have different lengths, with x y position of centroids. To classify them (supervised ...
1
vote
0answers
17 views

What types of transformation besides logarithm can I use for a linear regression?

I'm used to use log transformation in linear regression, most of times so I can get a normally distributed variable (even though I know this is not a requirement). I was wondering, what other ...
0
votes
0answers
11 views

Does Outliers in Categorical Feature varibales exist?

So, I was working on an Exploratory Data analysis project, after dealing with all the preprocessing of numerical features of the dataset when I started analyzing categorical data(nominal) there I ...
0
votes
0answers
18 views

Deep learning's auto extraction of representation

I am following up on a paper that demonstrates using deep learning (CNN) for classification. Specifically, their approach transformed the spatial data into fixed-length segments appropriate for CNN ...
1
vote
0answers
14 views

Get best performing feature

A data set that looks like this ...
2
votes
1answer
19 views

Why do we take the ratio of two features?

I've just learned that one technique for feature engineering consist of taking a ratio feature: feature1/feature2 , but I'm quite confused and I have questions: Why does this work ? Is this ...
0
votes
0answers
5 views

Cancel effect of external factor from dataset

In my problem I have a dataset with features: 1,2,3,...,n and another variable z, that is not part of my dataset, but every sample in the dataset has a corresponding z value. I might even see a trend ...
0
votes
1answer
20 views

sample space for a feature in machine learning [duplicate]

In machine learning data set each feature is considered as a random variable. Random variable is a function which maps the outcomes in sample space to a real value. Now I am trying to understand since ...
0
votes
1answer
24 views

Two ML models use different features. Does knowing the features of one model help improve the accuracy of the other model?

Suppose two firms are operating in the same field (e.g. insurance). If firm 1 knows which features firm 2 is using in their model, can firm 1 improve its model using that information? What if firm 1 ...
0
votes
0answers
13 views

Converting matrix to spatio-temporal feature vector

Given a matrix representing a video, where the rows represent video frames and the columns represent frame features, are there any techniques I can use to create a single feature vector from this ...
0
votes
1answer
45 views

How to discover relationships between features?

I'm an undergraduate fairly new to statistics. I am trying to find relationships between the features of a dataset. The dataset in question consists of 5000 objects each with up to 23 features. I was ...
0
votes
1answer
16 views

Does it make sense to do feature selection after PCA?

I have a dataset of 50 features that resulted after PCA was employed (originally, the dataset had 343 features. The 50 features are the principal components obtained with PCA). Does it make sense to ...
0
votes
0answers
29 views

How Helmert Encoding is done here? [duplicate]

I was going through an article on Helmert Coding there I encountered an example in which I was not able to understand how Helmert coding is done. I went through How to calculate Helmert Coding this ...
0
votes
0answers
10 views

Feature Engineering: How to deal with imbalanced numerical/categorical features

I'm analyzing a data set and solving a classification problem and find that values concentrate on one number in many features. For example, a categorical feature 'loan' indicating a person having loan ...
1
vote
0answers
38 views

About Feature Engineering Tips When “discriminative algorithm care about no modeling the probability of the language”

I was going over my old NLP course slides and one of the pages is about using Structured Perceptron for tagging. It claims that because the algorithm is discriminative, it doesn't care about modeling ...
1
vote
0answers
14 views

Classification of unnecessary pauses in a telephone call

So I'm working on a project where I'm interested in classifying a call as having excessive unnecessary pauses. This call has two speakers. And I also have the timestamps for each dialogue uttered by ...
1
vote
0answers
29 views

what is the impact of using large number of features compared to a small data?

I have time series data with 25 features and 181 observations, and the number of classes are 7. I used 3 models svm,random forest and xgboost to classify but the performance was really bad for each. i ...
1
vote
1answer
16 views

Using POS Tags and NERs as Features for Text Classification or Sentiment Analysis

I am trying to implement text classification and sentiment analysis from the documents. I always use POS tags as features in the following way. Mike is playing football I would convert it into ...
1
vote
0answers
6 views

Feature hashing and interpretability

I am working on a classification project where nearly all raw features (~30) are categorical such as category ID, city, or responsible team ID, some with cardinality > 600. I am using a feature hasher ...
0
votes
1answer
11 views

How to apply imputation when creating an analytical base table?

I am asked to read up on how to deal with missing values. From what I read I can use imputation with a package like MICE (for R) to automate this process. However I also read that when I am missing ...
0
votes
0answers
18 views

How to choose values for a Numeric Feature in a Decision Tree

I have a question about Random Forest and when it comes to Features for the Decision Tree. Let us assume that we have a Numeric Feature named: RSI Now let us assume that we have registered values ...
1
vote
1answer
20 views

Feature Engineering with Focus on KNN

I have seen a number of helpful posts such as this one on feature engineering, but I am specifically looking for something that may be helpful when using KNN. In my experience, some features work best ...
3
votes
1answer
34 views

Why do I get a different RSE when I call the summary of two identical linear models?

I am trying to find the best transformations for my data to use in a linear model. I call summary(lm(log10(Y)~X3^2) and get a RSE of 0.1787. Then I saved my data ...
0
votes
0answers
5 views

interesting features out of list of hair colors per person

lets say I have a dataset where each row -each data point - is a different person (about 1500 people). for each of them, I have several features. one of the feature is "hair color thought life". this ...
0
votes
0answers
10 views

Lasso Feature Selection on Batches of Training Data

I would like to perform lasso regression to select important features in a very large matrix (165K x 32K). I want to reduce the features from 32K prior to modeling, however, the file is too large to ...
0
votes
0answers
14 views

embedding a graph in n dimensional vector space and feed into a machine learning model

So I want to do some graph classification and regression. That is right! the training set are graphs. So my question is how can we embed a graph into a n dimensional vector space? Suppose the graph I'...
0
votes
0answers
12 views

How to control for primary predictive variable?

I'm trying to fit a model to predict customer churn. Specifically, I'm interested in how product feature adoption (e.g. web, mobile, digital services offered) affect churn. However, based on ...
1
vote
1answer
11 views

Multivalued feature construction

I have a task of representing a users feature matrix to be input for statistical machine learning classifiers, i have features like gender , age etc but I also have a multivalue feature called as "...
0
votes
0answers
9 views

What is the relationship (if any) between the number of input features and underfitting/ overfitting in ML/DL models?

Firstly below are a few points on this topic based on my understanding: Overfitting occurs when a particular model is too complex for the given data. This results in the model memorizing the data ...
0
votes
0answers
10 views

Continuous/Discrete - Categorical dependency test

I'm preparing the data to train a binary classification model. I know that eliminating features that do not have a dependency relationship with the target is a good practice. In my set of features I ...
1
vote
0answers
43 views

How to calculate Mahalanobis distance in very high dimensions with both continuous and categorical variables?

The objective is outlier detection via a distance measure. Does mahalanobis distance suffer from curse of dimensionality just like Eucledian distance for very high dimensions? (say around 5000 ...
0
votes
1answer
22 views

Fairly comparing performance of ML models trained with different numbers of features

I am training an ML model for a supervised classification problem. For this, I have been provided with two datasets. They both contain the same number of samples, however dataset 1 includes data for ...
1
vote
1answer
39 views

Expected Counts in Chi-Squared Goodness-of-Fit Tests of Normality

I have a variable with of 200 values that I would like to test for normality using the Chi-square Goodness of Fit test. To do this, I have to calculate, for each value, the expected value in a normal ...
0
votes
0answers
8 views

Priority between feature engineering and normalisation

My question is related to the priority between feature engineering (for example a simple transformation) and normalisation. It is a general question and I am not sure I understand all the ...
0
votes
0answers
7 views

Feature extraction for exponentially damped signals

I am looking into exponentially damped signals where it is a stationary signal (after implementing the Adfuller statistical test) and I would like to look into how can I extract meaningful features ...
0
votes
0answers
7 views

Is there I guide to decide which transformation to choose for different scenarios/ types of data and distribution?

1) how do i decide which transformation or scaling to use before passing our data into machine learning model. Can someone please guide me on which transformation to use in different situations. There ...

1
2 3 4 5
12