DEV Community

Model Evaluation Series' Articles

Back to Sachin Kr. Rajput's Series
Train/Validation/Test Split: Why Your Model Needs Practice, Dress Rehearsals, AND Opening Night
Cover image for Train/Validation/Test Split: Why Your Model Needs Practice, Dress Rehearsals, AND Opening Night

Train/Validation/Test Split: Why Your Model Needs Practice, Dress Rehearsals, AND Opening Night

Comments
10 min read
Accuracy, Precision, Recall, F1: The Four Judges Who Disagree on What Makes a Good Wolf Detector
Cover image for Accuracy, Precision, Recall, F1: The Four Judges Who Disagree on What Makes a Good Wolf Detector

Accuracy, Precision, Recall, F1: The Four Judges Who Disagree on What Makes a Good Wolf Detector

Comments
10 min read
When Accuracy Is a Lying Metric: The Weather Forecaster Who Was 96% Accurate and Still Got Everyone Killed
Cover image for When Accuracy Is a Lying Metric: The Weather Forecaster Who Was 96% Accurate and Still Got Everyone Killed

When Accuracy Is a Lying Metric: The Weather Forecaster Who Was 96% Accurate and Still Got Everyone Killed

Comments
10 min read
The Confusion Matrix: A Courtroom Drama Where Every Verdict Falls Into One of Four Boxes
Cover image for The Confusion Matrix: A Courtroom Drama Where Every Verdict Falls Into One of Four Boxes

The Confusion Matrix: A Courtroom Drama Where Every Verdict Falls Into One of Four Boxes

Comments
11 min read
AUC-ROC Explained: The Smoke Detector With a Sensitivity Dial and the One Number That Tells You If It's Any Good
Cover image for AUC-ROC Explained: The Smoke Detector With a Sensitivity Dial and the One Number That Tells You If It's Any Good

AUC-ROC Explained: The Smoke Detector With a Sensitivity Dial and the One Number That Tells You If It's Any Good

Comments
11 min read
Log Loss Explained: The Game Show Where Confidence Costs You — Being Wrong Is Bad, Being CONFIDENTLY Wrong Is Catastrophic
Cover image for Log Loss Explained: The Game Show Where Confidence Costs You — Being Wrong Is Bad, Being CONFIDENTLY Wrong Is Catastrophic

Log Loss Explained: The Game Show Where Confidence Costs You — Being Wrong Is Bad, Being CONFIDENTLY Wrong Is Catastrophic

Comments
12 min read
R-Squared Explained: The Dart Player Who's Somehow WORSE Than Just Aiming at the Center
Cover image for R-Squared Explained: The Dart Player Who's Somehow WORSE Than Just Aiming at the Center

R-Squared Explained: The Dart Player Who's Somehow WORSE Than Just Aiming at the Center

Comments
11 min read
MAE vs MSE vs RMSE: Three Bosses With Very Different Philosophies on Punishing Late Employees
Cover image for MAE vs MSE vs RMSE: Three Bosses With Very Different Philosophies on Punishing Late Employees

MAE vs MSE vs RMSE: Three Bosses With Very Different Philosophies on Punishing Late Employees

Comments
10 min read
Type I vs Type II Errors: The Fire Alarm That Cried Wolf vs The Fire Alarm That Slept Through Arson
Cover image for Type I vs Type II Errors: The Fire Alarm That Cried Wolf vs The Fire Alarm That Slept Through Arson

Type I vs Type II Errors: The Fire Alarm That Cried Wolf vs The Fire Alarm That Slept Through Arson

Comments
11 min read
Choosing the Right Metric: The Restaurant Inspector Who Judged Every Kitchen by Decor
Cover image for Choosing the Right Metric: The Restaurant Inspector Who Judged Every Kitchen by Decor

Choosing the Right Metric: The Restaurant Inspector Who Judged Every Kitchen by Decor

Comments
9 min read
Stratified Sampling: The Pollster Who Predicted a Landslide by Accidentally Surveying Only One Neighborhood
Cover image for Stratified Sampling: The Pollster Who Predicted a Landslide by Accidentally Surveying Only One Neighborhood

Stratified Sampling: The Pollster Who Predicted a Landslide by Accidentally Surveying Only One Neighborhood

Comments
11 min read
K-Fold Cross-Validation: The Comedian Who Tested Jokes at Only One Comedy Club and Bombed Everywhere Else
Cover image for K-Fold Cross-Validation: The Comedian Who Tested Jokes at Only One Comedy Club and Bombed Everywhere Else

K-Fold Cross-Validation: The Comedian Who Tested Jokes at Only One Comedy Club and Bombed Everywhere Else

Comments
13 min read
Validation Set vs Test Set: The Student Who Aced Every Practice SAT But Bombed the Real One
Cover image for Validation Set vs Test Set: The Student Who Aced Every Practice SAT But Bombed the Real One

Validation Set vs Test Set: The Student Who Aced Every Practice SAT But Bombed the Real One

Comments
10 min read
Is Your Model Good Enough to Deploy? The Restaurant That Served a 95% Perfect Dish — But the 5% Was Food Poisoning
Cover image for Is Your Model Good Enough to Deploy? The Restaurant That Served a 95% Perfect Dish — But the 5% Was Food Poisoning

Is Your Model Good Enough to Deploy? The Restaurant That Served a 95% Perfect Dish — But the 5% Was Food Poisoning

Comments
13 min read