Desktop · AcademicData

Sentiment Analysis

A desktop ML app that predicts sentiment from user text in real time, with an interactive retraining feature to improve the model on the fly.

Highlights

  • Binary classifier using ML.NET's SDCA Logistic Regression with automated text featurization.
  • Live retraining: users label new samples and instantly compare pre/post-training performance.
  • Evaluated with Accuracy, F1 Score, AUC, and LogLoss, all surfaced in the Windows Forms UI.

Tech Stack

C#ML.NETWindows FormsBinary Classification