Questions tagged [feature-selection]

Methods and principles of selecting a subset of attributes for use in further modelling

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

Find the most relevant columns for each single class in pandas

The following question (this one) did not help me. I have a big dataset, and I want to know which Columns are the most relevant for the Target Variable. I know that, in my case, for each class in the ...
0
votes
0answers
16 views

Find out which attribute of a movie causes the most variation in score

I'm tinkering around with a subset of the IMDb dataset, and have been thinking about what specific attributes of a movie impact its user rating. I am looking to survey what methods can be used to find ...
0
votes
0answers
6 views

Recursive feature elimination on train data or complete dataset and dummy encoding

I am using RFE with logistic regression. I will also be doing cross validation with RFE (RFECV in sklearn) to get the optimum number of features. I am not sure whether to use RFECV on just train ...
0
votes
0answers
24 views

if I got feature importance of xgboost/LightGBM what is next?

If I have feature importances of different variables in a xgboost/LightGBM model, how do I use this information? Is it better to just use the top n features and retrain the model? Does the feature ...
1
vote
1answer
17 views

My models performs better with the arbitrary random feature. How can I interpret this?

I am training 6 different classifiers 'Decision Tree', 'Random Forest', 'Logistic regression' and 'SVM' with different kernels. There are about 80 dependent variables including categorical and ...
0
votes
1answer
19 views

How to choose the features for an algorithm from the given attached screenshot?

How to choose the features from the given attached heat map & correlation factor for the classification algorithm? I have 6 different features i.e., ac233fc01403, ac233fc02eaa, ac233fc015f6, ...
1
vote
0answers
21 views

Increase accuracy of occupancy prediction?

I have a project that's aimed to predict the amount of occupants at my local gym given the date and weather. Here's my Kaggle kernel I have two datasets, occupants on a given hour and weather on a ...
1
vote
1answer
20 views

Implicit feature selection

I have heard that Random Forest and other tree based machines apply some kind of implicit feature selection. My Question is: Does this also apply for machines like the SVM? As far as I understand is ...
0
votes
2answers
22 views

How to deal with a feature that has lot of categorical values?

I know this question has been asked before and I have tried a few things but those things are not working as expected for my usecase. I have a 500 length feature vector. One of these features is a ...
2
votes
1answer
19 views

(Feature selection) In which cases it is legitimate to remove features manually?

I am dealing with the feature in which only one category takes up about 90%, the instances of more than 30 other categories are sparse. Is it reasonable to remove this feature before building an ...
1
vote
1answer
32 views

(Feature Selection) different results from L2-based and Tree-based

I am doing feature selection using Sklearn: Tree-based feature selection : RandomForestClassifier.feature_importances_ L2-based feature selection: LogisticRegression.coef_ Target variable is binary ...
2
votes
2answers
81 views

What do you call a feature that always has the same value?

Is there a standard term for a feature that always has the same value, i.e. that can be discarded without loss of information? For example I am trying to classify cats vs dogs, and every example in ...
2
votes
1answer
58 views

Feature selection is not that useful?

I've been doing a few DataScience competitions now, and i'm noticing something quite odd and frustrating to me. Why is it frustrating? Because , in theory, when you read about datascience it's all ...
1
vote
1answer
54 views

How to install boruta in conda?

I want to install boruta in my anaconda environment, but if I execute conda install boruta It displays, ...
1
vote
2answers
48 views

Improving classifcation when some are less represented?

I have a multi-class classification problem. It performs quite well but on the least represented classes it doesn't. Indeed, here is the distribution : And here are the classification results (I took ...
3
votes
0answers
25 views

Features selection with a lot of dummy variables in R

I am performing features selection on 3849 dummy variable (one-hot encoding) using Boruta algorithm and the algorithm is taking forever to run. Is there a faster way I can perform features selection ...
0
votes
1answer
15 views

Pearson correlation make disapear the target column

I have a dataframe with some numerical and categorical values. I want to do some feature selection to visualise a low-dimensional split in the dataset when the target variable is the grade. Yet, when ...
2
votes
1answer
87 views

Feature Importance

I have a dataset with 10 features. I've computed the feature importance using permutation importance with cross-validation from eli5, after fitting an extremely randomized trees (ET) classifier form ...
4
votes
0answers
53 views

Features reduction for the not correlated data set

I am working with classification problem on a training data set, which have 100 features. All the features in pairs haven't visible correlation. One can see it in the example pair plot for the some of ...
2
votes
1answer
25 views

Feature extraction in audio spectrogram

I have audio and its spectrogram of the words in English language. (A spectrogram is a frequency domain representation of a signal) Consider the words: chain, change, chair, chapter. As you can notice,...
0
votes
1answer
30 views

Feature selection, is it possible to combine wrapper and embedded methods?

I'm using neural network to predict PM10 concentration (a regression problem). Since the wrapper method is dependent on the model, so passing the neural network model that's optimized for all the ...
0
votes
1answer
46 views

xgboost feature selection and feature importance

when you create the new feature for data analysis for linear regression, it is clear that the feature has to be linear with other features is better but for xgboost what is the guideline to make a ...
1
vote
0answers
16 views

Difference between RFECV and SFS?

I used scikit.REFCV and mlxtend.SFS (backward) on the same data, same classifier, same cv, same scorer,... I also did a third version with sample weights passed to SFS's estimator And i'm conflicted ...
0
votes
1answer
39 views

What variance threshold to consider in feature selection?

Consider a numerical dataset with continuous variables, that has been scaled to end up with values in the [0,1] range. How can I compute a reasonable variance threshold for all the variables?
1
vote
0answers
21 views

Scikit-compatible network lasso implementation

Is anyone aware of a scikit-compatible network Lasso (nLasso) implementation? These papers have source code as well: D. Hallac, J. Leskovec, and S. Boyd, “Network lasso: Clustering and ...
0
votes
0answers
42 views

Feature selection for string columns

I tried sklearn SelectKBest for feature selection (feature scores) of supervised machine learning before. It looks good, but can only take numbers as inputs. Although I can present some string ...
1
vote
0answers
80 views

Explanation of how DeepExplainer works to obtain SHAP values in simple terms

I have been using DeepExplainer (DE) to obtain the approximate SHAP values for my MLP model. I am following https://github.com/slundberg/shap and DE's performance is very high in terms of computation ...
1
vote
2answers
54 views

Finding Feature Importance in CNN's?

Let's say I have images of cars. For each image in the dataset, I have let's say 3 pictures of the same car but in different angles. 1) The first image is the picture of the car from the front. 2) ...
0
votes
0answers
23 views

What is the dimension reduction method to large numbers of independent features while only two of them are important?why?

What is the dimension reduction method to model a data with large numbers of independent features (for instance 5k features), while only two of them are important (are effective in cost function)? I ...
0
votes
0answers
19 views

Multi-level data, what is the best approach?

I'm working on a dataset and having some problems. I hope you can give me your insight. So my objective is to predict customer churn based on incidents. Each incident is related to a contract which ...
3
votes
2answers
40 views

Decision Trees Should We Discard Low Importance Features?

I just started to work with feature selection. Let's say I have a decision tree model. I get its feature importances by tree.feature_importances_. In my model out ...
2
votes
3answers
80 views

Can ridge regression be used for feature selection?

I'm trying to figure out whether using Ridge Regression for regularization can be used to cause a more sparse hypothesis however to me it seems like ridge will never actually bring any coefficients to ...
1
vote
2answers
33 views

Feature selection filter methods

I am confused about when to use which filter methods for feature selection. I tried to learn them through online resources and found methods like chi-square, variance threshold, F-test, Mutual ...
2
votes
1answer
84 views

SelectKBest and Correlation returns me excatly same feature selection. How?

Im working on selecting most effective features from a dataset with over that 2000 features. Im using different algorithms for that (selectKBest with chi-square, Extra Trees, Correlation etc.) But ...
2
votes
2answers
90 views

Using random forest for selecting variables returns the entire dataframe

I am in the process of dimensionality reduction. I am using Random Forest to find the columns with the highest level of correlation with the target SalePrice column. The problem is that the output ...
0
votes
1answer
39 views

How to input a 3d model into ML algorithm?

I have a machine learning model that uses csv with measured data about buildings: width, length, height etc. I use it to predict some features and it works properly. I would like to drop csv with ...
1
vote
1answer
93 views

Light GBM Regressor, L1 & L2 Regularization and Feature Importances

I want to know how L1 & L2 regularization works in Light GBM and how to interpret the feature importances. Scenario is: I used LGBM Regressor with RandomizedSearchCV (cv=3, iterations=50) on a ...
0
votes
0answers
19 views

Feature selection for circular data in time-series

I'm predicting ozone concentration based on meteorological and seasonal variables. In the feature engineering stage I converted the MONTH, DAY_OF_WEEK, DAY_OF_YEAR to its sin and cosine components ...
2
votes
1answer
79 views

How does SHAP values help us to determine importance of a feature for a model trained by gradient boost?

I've read http://papers.nips.cc/paper/7062-a-unified-approach-to-interpreting-model-predictions.pdf and https://medium.com/@gabrieltseng/interpreting-complex-models-with-shap-values-1c187db6ec83 which ...
0
votes
1answer
14 views

Multivariate Multilag Regression with one shot prediction using LSTM

I am working on a multivariate regression task using a LSTM and I am interested in one shot prediction of my target variable (which is the price of a commodity). For example, the first parameter I ...
0
votes
1answer
28 views

Feature selection before or after applying filter in Time-series forecasting

I'm predicting ozone concentration based on meteorological variables and ozone value of the previous day. I applied savitzky golay filter to get rid of noise in the time-series dataset. My question ...
2
votes
1answer
35 views

Can anyone explain me the fisher score working

I have been working on feature selection and I wanted to know what does fisher score tell us about the data which helps us in feature selection.
0
votes
0answers
12 views

What are the common feature extraction technic to compare 2 sequence of timestamps?

I am building some predictive models for an online shopping site. I have timestamp log of customers' arrival time, time spent on a product's webpage, purchased or not, and a few others. I have tried ...
0
votes
1answer
13 views

How to include both origin and destination in your features?

I'm trying to predict the price of transportation for trucking freight. Two important features that I think would be of great impact are Origin and Destination. What's the best way to include that in ...
1
vote
1answer
26 views

For a regression model, can you transform all your features to linear to make a better prediction?

I was thinking. Would it be a good approach to check your features one by one (assuming you have a manageable amount of them) and see the relationship they have with your target variable, if they ...
3
votes
1answer
30 views

CrossValidation using glmnet and very high values of Lambda?

I am trying to run crossvalidation (folds=10) using glmnet library on my dataset. My outcome of interest is BMI and predictors include a set of clinical variables. My final goal is to use elastic-net ...
2
votes
2answers
72 views

unimportant features impact on model's performance

Using XGBoost and RandomForests, do unimportant features (according to the feature_importances_ attribute) hurt the model's performance? Do I need to carefully ...
1
vote
1answer
30 views

Using of 100s of Binary features in regression model

I have 100s of columns with binary values [0, 1] plus some extra columns without binary values. I am trying to do regression model but the model performance is very low. For non-binary features, I ...
0
votes
0answers
14 views

When we use VarianceThreshold from sklearn and do the transform so does our data also get scaled

I was working with Variance Threshold and when I used the transform function ,I found that the output is in floating points where as in original data set it was in integer format so why does it happen....
0
votes
0answers
24 views

records with perfect correlation to the answer. Drop or Keep?

I have about 1000 records (5 numeric, 5 categorical vars) and about 25 of them have something in 5-level categorical variable that just gives away answer. It's just too obvious and I'm not worried ...