Skip to main content

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
0 answers
7 views

Can manual feature extraction be considered a part of a learning algorithm?

We can view a learning algorithm as a tuple $(\mathcal{H}, \mathcal{O}, \mathcal{L})$ where $\mathcal{H}$, $\mathcal{O}$ and $\mathcal{L}$ are the hypothesis class, optimizer and loss function ...
ado sar's user avatar
  • 471
0 votes
0 answers
33 views

OLS - How to create a good feature

I'm working on trying to predict the next price return for a stock given the inside bid-ask volumes (level 1 data) and I want to create a feature that models the imbalance between bid volumes and ask ...
Joe's user avatar
  • 271
0 votes
0 answers
76 views

Is it bad not to standardize all features (regression)?

I'm working with a neural network with two hidden layers for a regression task. My output values for the training set vary from 0 to 2000 and for the test set from 0 to 600. My main problem is ...
stella's user avatar
  • 1
1 vote
1 answer
33 views

How can I find what is driving the change between my updated model and the original model?

I'm working on updating a model of rents (which is currently simple OLS) for my employer who has a large national portfolio. By tweaking here and there and drawing in a large amount of exogenous data ...
wjb_hwe's user avatar
  • 11
0 votes
0 answers
31 views

Could the community provide critiques on my method for picking features for modeling?

I just started a new job that is a bit more stats heavy than my old DS job. In my old gig I was mainly in the domain of scraping, descriptive statistics, visualizations, dashboards, etc. So I'm having ...
Nye307's user avatar
  • 11
0 votes
0 answers
17 views

Stock clustering based on fundamental reporting

I want to made a stock clusterization, based on their fundamental features from companies quarter reports. I collected quarter reports from 2018 to 2022. Some companies have reports for all quarters ...
TImur Nazarov's user avatar
0 votes
1 answer
24 views

can i have multiple similar features derived from same properties in a dataset?

Say I am fitting a (linear) regression model to a hundred-row dataset, whose data records a series of experiments that I can hardly reproduce or further conduct to get more records, I notice that ...
Yuuya's user avatar
  • 9
0 votes
0 answers
17 views

Target variable directly dependent on feature

Suppose a target of interest A and a feature B which serves as a good predictor for A. (both continuous variables) In literature, the target is always reported as C := A/B however. For me it is not ...
dinaue's user avatar
  • 1
0 votes
0 answers
6 views

Classification based on frequency decomposition of timeseries

I'm working on a classification problem where the dataset comprises a quote-unquote frequency profile from a timeseries. My dataset looks like this: ...
hernandezurbina's user avatar
0 votes
1 answer
46 views

Feature selection using backward feature selection in scikit-learn and PCA

I have calculated the scores of all the columns in my dataframe, which has 312 columns and 650 rows, using PCA. I used the following code: ...
Mostafa Bouzari's user avatar
0 votes
1 answer
44 views

How to Handle Infinite Values in Feature Engineering for Machine Learning Models

I'm currently working on a machine learning project where I am creating new features related to the ratio of bytes sent and received in a communications network. However, I'm facing a challenge: when ...
Camilo Piñón's user avatar
0 votes
0 answers
66 views

Understanding the Normalization/Standardization of geospatial coordinates

I'm building a neural network to predict future [latitude,longitude,altitude], and am having trouble dealing with the features. I've reviewed the answers to the ...
LivelyECDSA's user avatar
1 vote
0 answers
55 views

Riemannian alignment has no effect

I am trying to implement a Brain-Computer Interface system that should be able to differentiate between rest and movement trials. I am using motor-related cortical potentials for movement detection. ...
sjaustirni's user avatar
0 votes
0 answers
11 views

Should unprocessed features be kept in the dataset along engineered ones?

I'm working on a Machine Learning classification problem that has five Service Spending features (among many others) in its dataset, each sample is a customer. For instance, here are the first three ...
spengler's user avatar
1 vote
0 answers
34 views

How do i prioritize which features to use in my machine learning model before the feature engineering stage?

I am encountering a probably fairly common problem where I have too many features, lets say 500 possible features. I only want to pick the top 10-50 features that would be the most predictive of y, or ...
Katsu's user avatar
  • 991
8 votes
1 answer
326 views

How much does an inability to predict an apparent anomaly mean that we lack something in the feature space to distinguish it from business as usual?

I have read a number of questions where the crux is a lamentation that a rare outcome is unable to be predicted by a regression model of some kind. While I understand the desire to be able to reliaby ...
Dave's user avatar
  • 64.2k
1 vote
1 answer
34 views

Target variable is defined by combination of input features

I am trying to create a classification model which predicts whether or not a customer comes back to make a second transaction (after having made an initial transaction). I have details on date of ...
piper180's user avatar
  • 153
2 votes
1 answer
49 views

Adding a New Feature

My question is pretty straightforward and the task behind is related to binary classification. To add a new feature, do i first do train_test_split then add a new ...
Newbie's user avatar
  • 21
0 votes
0 answers
11 views

Toy dataset: Radial VAE

I'm evaluating disentanglement in toy datasets seeing as we have such little understanding of the phenomena. I'm using various tools from differential geometry. Now I want to train a VAE on the ...
John Miller's user avatar
4 votes
1 answer
125 views

Can XGBoost learn more complicated interactions/features?

For a set of features {a, b, c, d . . . n}, XGBoost can easily learn, say, a*d. In practice can it also effectively learn a/c? Or (a + b + c + 2)/d? Or (c^(2d))/(b^a)? I'd imagine some of this depends ...
BigMistake's user avatar
0 votes
1 answer
44 views

Questions about adding polynomial features to a dataset for linear regression

Apologies if this belongs to Data Science instead of here (I can move the question) but this seems related to the math aspect more than ML. In our course we just ...
evilmandarine's user avatar
0 votes
0 answers
13 views

Optimal method for predicting outcome from additive, correlated, and sparse features?

Suppose I have many vectors which can take on any of three values, 0, 1, 2. These vectors affect an outcome being predicted, Y. Vectors add together: a vector "A" of the value 2 has twice ...
BigMistake's user avatar
0 votes
0 answers
25 views

How to transform ratio input features in deep learning

I am a recommender system study which predict how likely a user browsing a product A will but another product B. One of the features is the price ratio of A and B, i.e., PriceB/PriceA. The assumption ...
Munichong's user avatar
  • 2,095
0 votes
1 answer
32 views

Working with subsets of values from single category in XGBoost

Since version 1.5, XGBoost supports categorical data out of the box, which is a convenient way to skip the one-hot pre-processing step and allow for if X in values ...
Alexandru Dinu's user avatar
0 votes
0 answers
17 views

Feature Engineering with Standardized Data

I'm relatively new to feature engineering, but at a high-level what I understand that it does is takes various features in a dataset (perhaps two, perhaps more) and combines them in such a way using ...
Hau5's user avatar
  • 21
0 votes
0 answers
37 views

Should you assign „0“ to a count variable that is not applicable?

I am designing a retrospective study in the field of oral and mandibular reconstruction. I am struggling with the interdependence of the following two variables and would greatly appreciate any input ...
Philipp's user avatar
0 votes
0 answers
15 views

Feature extraction methods for a distribution as exogenous variable in regression models

I have age distributions for more than $m>100$ different populations, each of varying sizes $(n_1, n_2, \dots, n_m)$. I'd like to create a regression model where these age distributions are used as ...
Girigio's user avatar
  • 63
1 vote
0 answers
25 views

What statistical test to use when comparing classifier performance for original dataset and new dataset with org + 1 variable

I would like to compare the classifier performance when there is a newly added variable to the dataset. Say the original prediction was with 10 input variables. The new one is with 11 inputs. My ...
mezbaha's user avatar
  • 11
1 vote
1 answer
13 views

Google android-app's titles and soft clustering

I have a no-trial question: I want to soft cluster the apps from Google Store. Most of the parameters are numbers, so no big clue. There are also "tags" but this is like using categorical ...
ozw1z5rd's user avatar
  • 171
0 votes
0 answers
32 views

Normalisation in feature extraction using pre-trained model

I have a dataset with medical images. I want to implement a network for super-resolution using GANs. One of the criteria of the optimisation is a perceptual loss. For that I will use a pretrained vgg ...
Janikas's user avatar
  • 101
1 vote
1 answer
69 views

How can I normalize features while preserving information about the original values?

I am trying to feed a neural network information about a stock (100s of features concerning price, MA, volume etc.). To ensure training stability, I normalize the features to have ...
Quantum's user avatar
  • 281
1 vote
0 answers
63 views

Should I split my dataset if I'm solely trying to understand feature importance?

I'm being provided a dataset with several variables in it, and a success metric (1 or 0) at the end. I'm being asked to analyze the dataset and give insights on how to improve the success metric rate....
Aradyan's user avatar
  • 11
1 vote
0 answers
24 views

Do similar PCA feature importance in first few top PCs mean these variables are nearly same in the original space?

I am using PCA to do the data inspection. First 3 PCs explain nearly 82% of the total variance. Suppose the number of features is $n$. And I found 4 of the features have similar PCA feature ...
Xu Shan's user avatar
  • 163
2 votes
1 answer
86 views

Can variables used for rule based labeling be treated as input features?

I am currently working on binary classification problem with imbalanced dataset (n=3419 and 69:31). However, based on the business expertise of the users, they have generated rule-based label based on ...
The Great's user avatar
  • 3,302
0 votes
0 answers
60 views

Feature contribution interview question I can't answer

A few weeks ago, I had an interview for a data science job. Of all the questions they asked me, I was unable to solve the following one. I couldn't even attempt it because I didn't know anything about ...
Girigio's user avatar
  • 63
1 vote
1 answer
1k views

Can we do sine , cosine , tan and cot transformation in regression?

Can we transform the variables of a regression in MLR to sine , cosine, tan Then how to interpret the results if I get a good $R^2$ and good adjusted $R^2$
sriram's user avatar
  • 59
0 votes
0 answers
13 views

What are the ways to use a LIST of features which are DYNAMIC (contents) in nature?

Any features which is represented as a list of 0 or more elements is what I call a Dynamic feature. Let us suppose an example where there are 10 Million movies and ...
Deshwal's user avatar
  • 244
6 votes
2 answers
355 views

Options for 3D coordinate systems?

I'm trying to solve biochemistry problems (think protein folding) with DNNs. Are there 2D / 3D coordinate systems that are particularly well suited for deep neural networks (DNNs) to process? For ...
Yaoshiang's user avatar
  • 271
2 votes
3 answers
579 views

What are the methods to increase the dimension of a feature space?

Is there a way to increase the number of dimensions through feature transformation in machine learning? If so what are the techniques involved?
4 votes
0 answers
107 views

When does target encoding lead to overfitting

Let us say, we are tasked with setting (average/list) prices that are likely to convert for heterogeneous products (e.g. used cars of all shape and sizes - made up example!). Let us also say that we ...
cs0815's user avatar
  • 2,245
1 vote
0 answers
41 views

Encoding ordinal categories as features

When we have categorical features in a regression (say a generalized linear model for now), it is typical to let one category be subsumed by the intercept and then code binary indicator variables for ...
Dave's user avatar
  • 64.2k
1 vote
0 answers
27 views

Should I destandardize the errors from training the neural network?

So, I am learning a bit about Neural Networks. I have built a code in PyTorch for regression, and I have standardized both the features and the target variables following this answer. My question is ...
No-Time-To-Day's user avatar
1 vote
1 answer
869 views

Why data-scaling in range (0,1) is important? [duplicate]

Often a preprocessing technique to do is to normalize our data in a range (0,1) before we tow our model (example neural network) on them. I understand why in a practical way (for example if we have ...
pietrus's user avatar
  • 11
2 votes
0 answers
35 views

Does it make sense to transform a feature containing hours (24h) into two features with xy-coordinates of each hour in the space? [duplicate]

I have a clustering problem that I might solve with an algorithm based on Euclidean distance (e.g. K-Means). One potential feature is the "hour" at which each user began an interaction. As ...
rusiano's user avatar
  • 566
1 vote
0 answers
136 views

Why cannot I use silhouette score with ground truth labels?

I was looking into checking cluster positioning from a non-liner transformation. I do have the ground truth labels. Now, I want to use the transformed data points and see how good this transformation ...
ponir's user avatar
  • 111
1 vote
1 answer
49 views

Why is not a scalar feature enough to encode 3-component binary numbers in an autoencoder?

I am trying to build an intuition on what really a feature is. I created a toy example as following. In my mind a scalar feature should be enough to represent my data. Couldn't the model in this case ...
ElPotac's user avatar
  • 51
1 vote
1 answer
358 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 ...
Lorenzo's user avatar
  • 11
2 votes
2 answers
74 views

Curve quantification

I have some longitudinal measurement data of 15,000. I smoothed that data with B-spline smoothing and got the following curve. I then want to quantify this curve and extract features for clustering ...
NakataKoo's user avatar
1 vote
1 answer
224 views

How to encode categorical variable with multiple categories per datapoint?

Consider this question on a survey: What desserts have you eaten? Apple pie Banana pudding Coconut cake Doughnut holes The user can pick as many of the options as they like. How would one encode ...
xojfqa's user avatar
  • 177
1 vote
1 answer
39 views

Having to use features which have low correlation with the target

I'm applying LogisticRegression on breastcancer dataset. Steps : - 1- A correlation matrix resulted in only four features having ...
Pixel_Bear's user avatar

1
2 3 4 5
16