Victor Alejandro Leiva Espinoza

Data Scientist - Portfolio

---

Project maintained by Vpy7 Hosted on GitHub Pages — Theme by mattgraham

📍 San Pedro de la Paz, Chile 📧 leivav177@gmail.com 💼 LinkedIn 🐙 GitHub

---

Project Portfolio

---

Education

Master’s Degree in Data Science for Innovation

Universidad de Concepción | Mar 2024 – Present

Bachelor’s Degree in Astronomy

Universidad de Concepción | Mar 2020 – Nov 2023

---

Experience

Data Scientist (Intern)

Centro de Datos e Inteligencia Artificial, Universidad de Concepción | Jan 2025 – Present

• Built a PPO-based RL system for Turtlebot3 robots in Webots, improving task success rate by 25% in 2 months.
• Simulated multi-robot behavior (up to 6) with Deepbots.
• Developed a Python dashboard to monitor training performance.

Data Analyst – AI Training

Telus International | Aug 2024 – Present

• Process and clean complex datasets to ensure quality for AI models.
• Work under NDA ensuring confidentiality of metrics and outputs.

Intern – Data & ML Workshops

Harvard John A. Paulson School of Engineering and Applied Sciences | Jun 2024

• Attended workshops on visualization, regression, classification, trees, and neural networks.
• Participated in innovation and startup talks and hands-on ML labs.

Teaching Assistant – Statistics

Departamento de Estadística, Universidad de Concepción | Aug – Dec 2023

• Delivered weekly lectures for Astronomy & Geology students.
• Covered descriptive statistics, distributions (Normal, Binomial, etc.), and inference tests (χ², t-test, Fisher).
• Improved course materials with astrophysical applications.

Research Assistant / Data Scientist

Núcleo Milenio TITANs | Feb – Dec 2023

• Built and tested Python pipeline with pPXF to estimate SMBH masses from 1300+ galactic spectra (LAMOST, SDSS).
• Queried/processed large datasets with SQL and remote servers via SSH.
• Reported weekly results to PI.

Intern

Núcleo Milenio TITANs | Jan 2023

• Contributed to early development of the pPXF pipeline in Python.
• Updated third-party code to Python 3.10 and pPXF 8.2.

---

Technical Skills

I include all relevant languages, packages and methodologies in which I have proven experience.

Programming Languages

Python

Libraries & Frameworks

Data Manipulation & Analysis

Pandas NumPy

Data Visualization

Matplotlib Seaborn Plotly

Machine Learning

Scikit-Learn XGBoost TensorFlow Keras PyTorch RandomForestClassifier SMOTE ADASYN GridSearchCV StratifiedKFold LabelEncoder StandardScaler ClassificationReport ROC-AUC Precision-Recall Curve Threshold Adjustment

Computer Vision

OpenCV Object Detection Image Classification

Reinforcement Learning

Deepbots

Time Series Analysis

ARIMA Statsmodels

Data Science Methodologies

Supervised Learning Unsupervised Learning Reinforcement Learning Hypothesis Testing Experiment Design Feature Engineering One-Hot Encoding Hyperparameter Tuning Monte Carlo Simulation

Math & Stats Foundations

Linear Algebra Calculus Statistics Probability

Databases & Querying

PostgreSQL MySQL SQLite SQL Data Modeling

APIs & Web Technologies

REST Web Scraping BeautifulSoup Selenium JSON YAML XML

Tools & Environments

Jupyter VSCode Google Colab GitHub SSH Google Sheets LaTeX Markdown

BI & Reporting Tools

Power BI

Data Handling Skills

Data Cleaning Data Visualization

---

Languages

• Spanish: Native
• English: Advanced

---

References

Dr. Neil Mark Nagar – Letter

• Director, Núcleo Milenio TITANs

  nagar@astro-udec.cl

María José Medina Fritz, M.Sc. – Letter

• Departamento de Estadística, Universidad de Concepción.

  mmedinaf@udec.cl