The official PyTorch implementation of the paper "SAITS: Self-Attention-based Imputation for Time Series". A fast and state-of-the-art (SOTA) deep-learning neural network model for efficient time-series imputation (impute multivariate incomplete time series containing NaN missing data/values with machine learning). https://arxiv.org/abs/2202.08516

Phasing and genotype Imputation comparison. Have been evaluated: BEAGLE 5.4, EAGLE 2.4.1, SHAPEIT 4, MINIMAC 4, IMPUTE 5, using accuracy metrics like: IQS(Imputation Quality score), r2 (Pearson correlation), Concordance.

Missing value imputation package in Python specialized for High-performance computing.

This my entry for the Titanic competition on Kaggle. May 2019: public score is 0.80382, which is a top 10% ranking on the leader board of around 11.249 participants.

Exploratory data analysis of geochemical and biogeochemical data for use in mineral exploration and environmental studies.

This package focuses on the tasks of dealing with outlier and missing values, scaling, and correlation visualization.

Predict song popularity using IterativeImputer, PCA, and Random Forest

Getting data ready for ML