A take on highly imbalanced fraud classification using permutation importance to select top features and explaining the model using SHAP.

Feature importance refers to a measure of how important each feature/variable is in a dataset to the target variable or the model performance. It can be used to understand the relationships between variables and can also be used for feature selection to optimize the performance of machine learning models.

Developed a machine learning model using scikit-learn, implementing ensemble techniques, PCA, correlation analysis, and extensive feature engineering. The goal was to classify documents as either human-generated (0) or AI-generated (1) based on document embeddings, word count, and punctuation.

Contains analysis of Lyft ride attributes and how it affects demand surge in the city of Boston.

A take on highly imbalanced fraud classification using permutation importance to select top features and explaining the model using SHAP.

Comparing different tree-based algorithms to find the best model for cancelation prediction

End-to-end project to analyze and model concrete compressive strength data then productionize the best model to help civil engineers determine concrete structural integrity

Klasifikasi Musik Berdasarkan Genre Menggunakan Metode Naive Bayes.

This repo is all about feature importance. Whereby we look at the ways one can identify if a feature is worth having in the model or rather if it has a significant influence in the prediction. The methods are model-agnostic.

High data dimensionality and irrelevant features can negatively impact the performance of machine learning algorithms. This repository implements the Permutation feature importance method to enhance the performance of some machine learning models by identifying the contribution of each feature used.

This project aims to study the influence factors of international students' mobility with the case of international students from B&R countries studying in China.

These training sessions in machine learning, conducted by Yandex, are dedicated to classical machine learning. This offers an opportunity to reinforce theoretical knowledge through practice on training tasks.

This project fits and tunes several regression models to predict Parkinson's symptom severity scores from voice recordings.

This project contains the data, code and results used in the paper title "On the relationship of novelty and value in digitalization patents: A machine learning approach".

A solid foundational understanding of XAI, primarily emphasizing how XAI methodologies can expose latent biases in datasets and reveal valuable insights.

eXplainable Artificial Intelligence (XAI) Basic Algorithms on Iris Dataset

A Python Package that computes Target Permutation Importances (Null Importances) of a machine learning model.