Dmitriy Bolotov

Welcome

I am a data scientist focused on applied machine learning and practical AI.

Over the past decade, I’ve worked in adtech and IoT & industrial analytics, developing solutions using predictive modeling, anomaly detection, and deep learning.

This site shares projects and open-source tools I've worked on.

Tools

Time Series Smoothing Visualizer - Interactive app that helps users understand six common smoothing techniques for time series data. | GitHub

Visual Time Series Generator - Interactive app for generating time series data. Includes options for missing values and anomalies. | GitHub

DS QuickRef - A concise, opinionated reference for practical data science and machine learning workflows in Python.

Python and R Packages

Python and R packages for exploratory analysis, model development, and deployment:

• minieda – Minimalist Python package for summarizing DataFrames.
• pmml – Export predictive models and preprocessing steps from R to PMML, an open standard for deployment.
• neighbr – K-nearest neighbors for classification, regression, and clustering in R.
• amelie – Anomaly detection using maximum likelihood estimation. | Intro

Case Studies

Predicting Gym Member Churn for Targeted Retention – Includes cohort breakdowns, predictive model, risk segmentation, and targeted retention strategies. | Medium article | Jupyter notebook