MACHINE LEARNING // MODEL INFERENCE // 10

Income Prediction ML App.

Built a Streamlit inference application that loads a saved scikit-learn pipeline and predicts an income class from demographic and employment inputs.

CLIENT

Academic ML Project

ROLE

Machine Learning Developer

CHALLENGE

Consistent inference requires the same data cleaning, feature transformation, and classifier configuration used during model training.

SOLUTION

Implemented a reusable pipeline with missing-value handling, categorical encoding, Gradient Boosting classification, serialized model artifacts, probability output, and model-evaluation panels.

KEY FEATURES

01Streamlit inference form
02Saved scikit-learn pipeline
03Missing-value preprocessing
04Categorical feature encoding
05Gradient Boosting classifier
06Class probability output
07Stored model metrics
08Reusable model artifact

TECHNOLOGY

Pythonscikit-learnPandasNumPyStreamlit

RESULTS

Delivered a deployed inference interface that connects validated user input to a serialized preprocessing and classification pipeline.

PROJECT LINKS

LIVE DEMOVISIT

GITHUBVIEW

NOTE

The prediction is a portfolio model output, not financial, employment, eligibility, or policy advice. It must not be used to make high-impact decisions.

NEXT PROJECT

Word Segmentation OpenCV.

COMPUTER VISION // IMAGE SEGMENTATION