Update README.md

README.md

@@ -40,7 +40,7 @@ pipeline_tag: tabular-classification
 
 ## Project Overview
 
-This project builds a complete end-to-end machine learning pipeline that predicts developer compensation using the Stack Overflow Developer Survey 2023 (~89,000 developers worldwide). The pipeline progresses from raw data through exploratory analysis, feature engineering, unsupervised clustering, regression modeling, and multi-class classification — ending with two production-ready Gradient Boosting models exported to this repository.
+This project builds a complete end-to-end machine learning pipeline that predicts developer compensation using the Stack Overflow Developer Survey 2023 — a dataset of ~89,000 developers worldwide with 84 raw features. From those, I selected 16 features most relevant to salary prediction and ended up with 45,804 developers after cleaning.
 
 The same dataset is used for two prediction tasks:
 - **Regression** — predicting the exact annual salary in USD.
@@ -337,7 +337,9 @@ print(f"Class probabilities: {dict(zip(clf_model.classes_, predicted_proba[0]))}
 ## Dataset
 
 **Source:** [Stack Overflow Developer Survey 2023](https://www.kaggle.com/datasets/stackoverflow/stack-overflow-2023-developers-survey) on Kaggle  
-**Original size:** ~89,000 respondents, 16 features used  
+**Original size:** ~89,000 respondents, 84 features in the raw dataset  
+**Selected for analysis:** 16 features chosen for relevance to salary prediction  
+**After cleaning:** 45,804 developers with valid salary data ($5K–$500K range)
 **After cleaning:** 45,804 developers with valid salary data
 
 ---