List: Data sci, ML, metrics, imbalanced, class probability, ROC... | Curated by David B Rosen (PhD)

Jan 11, 2023
41 stories
Data sci, ML, metrics, imbalanced, class probability, ROC...Interesting articles about machine learning, performance metrics, probability predictions, imbalanced classification AND regression, and more
Bozhong Liu
Three effective ways to deal with domain gap and imbalanced data in multi-class classificationHow to deal with domain gap and imbalanced data in a large dataset.
May 28, 2021
May 28, 2021
In
Heartbeat
by
Kris Ograbek
Cost-Sensitive Learning for Imbalanced DataTraining good models that deal with imbalanced data is challenging. Cost-sensitive training is one way of doing it.
Oct 8, 2021
Oct 8, 2021
Mridul Katta
Handling Imbalanced DatasetImbalanced datasets are normal in real life. This article talks about a few good strategies to create & evaluate the balance in the…
Aug 10, 2021
Aug 10, 2021
In
TDS Archive
by
Susan Currie Sivek, Ph.D.
Balancing Act: Classification with Imbalanced DataHelp your data find balance to make sure you create a quality classification model
Nov 4, 2021
Nov 4, 2021
In
TDS Archive
by
Samuele Mazzanti
You Can Compute ROC Curve Also for Regression ModelsThey probably told you that Area under ROC curve cannot be computed for a continuous target variable. Well, they were wrong. Here is how to…
Sep 16, 2021
5
Sep 16, 2021
5
In
TDS Archive
by
Samuele Mazzanti
Your Dataset Is Imbalanced? Do Nothing!Class imbalance is not a problem. Debunking one of the most widespread misconceptions in the ML community.
Aug 24, 2022
29
Aug 24, 2022
29
In
TDS Archive
by
Samuele Mazzanti
Data Scientists Need to Know Just One Statistical TestAfter you read this, you will be able to test any possible statistical hypothesis. With a unique algorithm.
Jun 30, 2022
21
Jun 30, 2022
21
In
TDS Archive
by
Scott Lundberg
Interpretable Machine Learning with XGBoostThis is a story about the danger of interpreting your machine learning model incorrectly, and the value of interpreting it correctly…
Apr 17, 2018
49
Apr 17, 2018
49
In
TDS Archive
by
Steven Dye
How to Handle SMOTE Data in Imbalanced Classification ProblemsKnow where the pitfalls are and how to avoid them
May 2, 2020
May 2, 2020
In
TDS Archive
by
Becca R
Imbalanced Class Sizes and Classification Models: A Cautionary Tale Part 2Recently, I wrote this post about imbalanced class sizes in classification models might lead to overestimation of a classification model’s…
Mar 27, 2019
5
Mar 27, 2019
5
In
TDS Archive
by
Kanellis Georgios
Proper Balancing for Cross ValidationWho hasn’t come across the need of applying the cross validation technique, while the dataset in hand is imbalanced, in regards to the…
Oct 11, 2019
2
Oct 11, 2019
2
In
TDS Archive
by
KSV Muralidhar
The right way of using SMOTE with Cross-validationThis article discusses the right way to use SMOTE to avoid inaccurate evaluation metrics while using cross-validation.
Mar 29, 2021
9
Mar 29, 2021
9
In
Analytics Vidhya
by
Ayobami Akiode
How to carry out k-fold cross-validation on an imbalanced classification problemAn imbalance classification has its OWN rules. Know them,  else you violate their rights.
Jul 10, 2020
2
Jul 10, 2020
2
David B Rosen (PhD)
Good article!For classification evaluation metrics, I think the first and foremost distinction that must be drawn among them is those that evaluate a…
Jun 28, 2021
Jun 28, 2021
David B Rosen (PhD)
There's an easier way to vectorize your example in pandas in just a single assignment of a single…Or alternatively you could nest two np.where() in a single assignment as follows, which is about twice as fast as above (about 1300 times…
Aug 24, 2021
1
Aug 24, 2021
1
In
Lumiata
by
Lumiata
Cross-Validation for Imbalanced DatasetsAnusha Mohan, Data Scientist
Mar 5, 2019
8
Mar 5, 2019
8
David B Rosen (PhD)
To @Panos V. ’s question, “ How to implement this with a GridSearchCV though?”:I’ve done it this way:
Jan 24, 2020
Jan 24, 2020
In
TDS Archive
by
Jonathan Grandperrin
How to use confidence scores in machine learning modelsAs humans, machine learning models sometimes make mistakes when predicting a value from an input data point. But also like humans, most…
Jan 19, 2021
6
Jan 19, 2021
6
David B Rosen (PhD)
R² (R-squared) is a very useful transformation of MSE which simply scales and offsets it relative…
Jun 28, 2021
Jun 28, 2021
In
TDS Archive
by
João Paulo Figueira
Stratified Splitting of Grouped Datasets Using OptimizationThis article explains how to perform a stratified split of a grouped dataset into train and validation sets.
Feb 23, 2021
Feb 23, 2021