Logistic Regression - Classification with C# ML .NET

preview_player
Показать описание
Training an ML model to predict the heart disease outcome. We'll see how Logistic Regression is different from Linear Regression before we dive into the model building. Afterwards, we'll go over the results & metrics using visualizations and more!

00:00 Intro
00:40 Theory
07:55 Exercises
09:45 Classification with ML.NET
16:40 Evaluate & Explain Metrics
23:45 Try Model & Conclusion
27:35 Outro

#classification #logistic #regression #linear #model #machinelearning #basics #theory #simplification #artificialintelligence #mathematics #statistics #dotnet #mlnet #modelbuilder #predictions #features #vector #tensor #data #datascience #dataanalytics #datastructures #csharp #csharptutorial #visualstudio #usecase #customeracquisition #forecasting #priceprediction #vscode #codefirst #mlapi #api #nuget #computervision #nlp
Рекомендации по теме
Комментарии
Автор

@23:11 I took out the positive & negative out of the definitions making it less clear.
- (Positive) Precision: the proportion of correctly predicted (positive) instances among all the (positive) predictions.
- (Negative) Precision: the proportion of correctly predicted (negative) instances among all the (negative) predictions.

kis.stupid
Автор

@26:45 After plotting the amount of females with heart disease, only 24 vs. 108 males with the disease, I can assume that this data imbalance makes that the model will predict "No disease" when given input data about a female. I will make a post about this.
Plotting the amount of people with the diseases by age, could've shown us a similar imbalance. Conclusion, for a well-performing model, we'd want an evenly-distributed / well-balanced dataset to train on.

kis.stupid
welcome to shbcf.ru