Skip to content
View nsdmlk's full-sized avatar
⚜️
Cooked
⚜️
Cooked

Block or report nsdmlk

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
nsdmlk/README.md
cover



between mathematics & machines

small data · strange models · quiet experiments


ML researcher and Python developer focused on small data and conformal prediction — models and libraries that stay honest when samples are few.

Currently at Beijing Institute of Technology, Computer Science.



Ⅰ · Projects

SELECTED WORKS · 2026



01 · SmallGBM

gradient boosting for small datasets

0.9101 ROC-AUC · 13.4 ms fit · 0.29 ms predict · 22 datasets



02 · SmallMLP

non-parametric regression & classification with conformal sets

NON-PARAMETRIC · CONFORMAL SETS · SMALL DATA



Ⅱ · RESEARCH

PREPRINTS · PUBLICATIONS



SmallGBM

Gradient Boosting with Robust Leaf Regularization for Small-Sample Tabular Data

ZENODO · 2026 · Emelyanov, I. · DOI: 10.5281/zenodo.21934674



SmallMLP

Non-Parametric Regression & Classification with Conformal Sets

PREPRINT · IN PREPARATION · Emelyanov, I. · PyPI · smallmlp



Ⅲ · Skills

TOOLS · DAILY DRIVERS



LANGUAGES   



ML / DATA   



BACKEND / TOOLS   





footer

☾ BEIJING · CHINA · 2026 ☾

Pinned Loading

  1. SmallGBM SmallGBM Public

    SmallGBM is a gradient boosting library designed for small datasets (n < 1000). It combines robust leaf weight estimation with stochastic split selection to outperform XGBoost and LightGBM — with l…

    Jupyter Notebook 10

  2. meme-mirror meme-mirror Public

    Real-time CV plugin that watches your camera, detects your emotion/pose, and retrieves the most similar meme via multimodal embeddings (CLIP + FAISS). Built with Python, MediaPipe, and FastAPI.

    Python

  3. SmallMLP SmallMLP Public

    SmallMLP is a library for small, nonlinear datasets ( n < 500 ). It replaces manual architecture search with adaptive hyperparameters derived from dataset parameters ( n , d , K ), and produces cal…

    Python 6

  4. sRFTM sRFTM Public

    A small transformer that converts informal math expressions into LaTeX.

    Jupyter Notebook 1

  5. SmallML SmallML Public

    Локальный AutoML-инструмент для малых данных (n < 1000). Фокус: научные лаборатории, исследователи, малые датасеты.

    1

  6. SmallGP SmallGP Public

    Gaussian Processes for small nonlinear data: adaptive kernels, sparse approximation, calibrated intervals.

    Python 1