Skip to content
View evandeilton's full-sized avatar
🏠
Working from home
🏠
Working from home
9 followers · 6 following

Achievements

Achievement: Pull Sharkx2Achievement: Pair ExtraordinaireAchievement: QuickdrawAchievement: Arctic Code Vault Contributor

Achievements

Achievement: Pull Sharkx2Achievement: Pair ExtraordinaireAchievement: QuickdrawAchievement: Arctic Code Vault Contributor

Block or report evandeilton

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
evandeilton/README.md

José Evandeilton Lopes

Data Scientist @ Stix · PhD Candidate in Statistics (PPGMNE/UFPR) · Credit Risk & Fraud Analytics

Profile Views LinkedIn GitHub ORCID

About

  • Data Science Specialist @ Stix, leading Advanced Analytics: ~30 ML models in production, fraud detection in point redemption, and dashboards for a 16M-member loyalty ecosystem
  • PhD Candidate in Statistics (PPGMNE/UFPR, advisor Wagner Hugo Bonat), researching MCGLM-Graph — extending Multivariate Covariance GLMs with graphical-model representations of residual dependence (GGM & DAG layers)
  • 10+ years in credit risk & fraud modeling across HSBC · Bradesco · ClearSale · Boa Vista SCPC
  • Registered Statistician (CONFE)

Research interests: Graphical Models · Bayesian Networks · Multivariate GLMs · Computational Statistics · Credit Risk & Fraud Modeling

Open-Source Packages


Generalized Kumaraswamy regression for bounded (0,1) data
Companion 7-family bounded distribution engine (C++/RcppArmadillo)
Beta regression for interval-censored scale outcomes · with W. Bonat
High-performance optimal binning & WoE for credit scoring

Also on GitHub: dscoder (AI code-generation agent for data science) · rnp (teaching toolkit behind the R na Prática book)

Skills

R Python SQL PySpark Julia C++ TMB Databricks Azure

Credit Scoring PD/LGD/EAD Fraud & AML Bayesian Statistics GLM/GAM/GLMM MCMC Survival & Multivariate Analysis Causal Inference XGBoost/LightGBM

Academic Background

  • Ph.D. in Statistics (in progress, PPGMNE/UFPR, 2024–2027) — Extensão de Modelos MCGLM: Uma Abordagem via Modelos Gráficos Probabilísticos (MCGLM-Graph) · advisor Wagner Hugo Bonat · pré-defesa concluída em 2026
  • M.Sc. in Statistics (PPGMNE/UFPR, 2021–2023) — Beta regression models for scale data
  • B.Sc. Statistics (UFPR, 2009–2013) · MBA Business Intelligence (Universidade Positivo, 2014–2015)

Let's Connect

Open to collaborations on statistical modeling, credit risk, graphical models, and R package development.

Email Google Scholar ORCID

Fun Facts

personal_attributes <- list(
  favorite_distribution = "(gkw) Generalized Kumaraswamy Distribution",
  life_motto = "In God Trust",
  superpower = "Finding patterns in chaos",
  currently = "Turning residual covariance matrices into graphs"
)

Popular repositories Loading

  1. OptimalBinningWoE OptimalBinningWoE Public

    OptimalBinningWoE is a high-performance R package for optimal binning and Weight of Evidence (WoE) transformation, designed for credit scoring, risk assessment, and predictive modeling applications.

    C++ 4

  2. rnp rnp Public

    Pacote R do projeto R NA PRÁTICA: estatística descritiva, probabilidade, inferência, regressão, multivariada e séries temporais, com rotinas em C++ (Rcpp/RcppArmadillo) e saídas tidy. Documentação …

    R 2

  3. gkwreg gkwreg Public

    gkwreg: Generalized Kumaraswamy Regression Models for Bounded Data. The package provides a comprehensive and computationally efficient framework for regression modeling of data restricted to the st…

    R 2

  4. scoreranking scoreranking Public

    Governed binary risk ranking and additive scorecard development for credit and fraud risk in R

    R 2

  5. betaregscale betaregscale Public

    The betaregscale package is a library of R functions for fitting beta regression models to interval-censored transformed scale data, allowing for fixed or variable dispersion.

    R 1

  6. gkwdist gkwdist Public

    gkwdist implements the Generalized Kumaraswamy (GKw) distribution family and its seven nested sub-models for bounded continuous data on ( 0 , 1 ) . All functions are implemented in C++ via RcppArma…

    R 1