ML hyperparameters tuning and features selection, using evolutionary algorithms.

[NPJ Com Mat 2023 | Small 2024] Machine Learning Algorithm : outlier identifying, feature selection

Particle Swarm Optimization (PSO) - Feature Selection and Hyperparameter Tuning

A smart Network Intrusion detection tool to perform forensics on your network.

This project predicts traffic accident severity using the 2023 STATS19 dataset. It employs machine learning and LIME to not only forecast outcomes like 'Fatal' or 'Slight' but also to provide clear, understandable explanations for its predictions, creating a powerful tool for enhancing road safety.

Comprehensive benchmark study of feature selection techniques for predictive machine learning models on tabular data. Various feature selection methods are evaluated across different data characteristics and predictive scenarios.

This Repository Contains Different Machine Learning and Important Concepts

主要包含ModelHelper和NLPHelper，其中ModelHelper主要有特征选择、超参数搜索、模型解释和模型融合等，NLPHelper则是进一步封装了NLP一些常用的操作，常用的网络结构以及几个NLP的任务

AlmaBetter Capstone Project - Unsupervised ML. Developed a book recommendation system for Amazon customers using memory and model based collaborative filtering by utilizing the description of book consumed and user interests.

End-to-end brain tumor segmentation on BraTS2020 with a modified DUCKNet (U-Net + DenseNet). Includes data preprocessing/augmentation, Keras training loops, and rigorous eval (Dice/IoU), achieving 88.7% validation Dice with 0.010 validation loss. Reproducible notebook and comparisons vs baseline U-Net; trained on A100.

Data Science - Neural Networks Work

FeatEngX is an automated Feature Engineering Tool used by Data Engineers & AI Researchers for making feature selection process, data preprocessing, and engineering accurate.

A machine learning project to predict bike rental demand using historical data. This project uses an end-to-end workflow including data preprocessing, feature engineering, and a tuned XGBoost Regressor model to achieve high predictive accuracy.

Forward and backward feature selection with adjusted R squared in R

EDA

This repository describes the methods used to test different sci-kit learn feature selection methods as part of Qiime2 q2-classifier.

Data Science - Random Forest Work

Data science project applying feature selection/dimensionality reduction techniques to identify the explanatory variables to be included within a linear regression model that predicts the number of times an online news article will be shared using Python 3 in a Juypter Notebook.

Giving a song dataset, thorough exploratory analysis, diverse model construction, and innovative feature engineering to developing predictive models for song scoring.