Alex Amaguaya

Research Assistant / Data Scientist

Biography

Alex works as a Data Scientist and usually collaborates in ESPOL research projects. He also has experience working with Social Media and Financial data. And he is interested in developing research involving Econometrics and Machine Learning (intersection of fields). Now, I’m a first year Econometrics master student at TU Dortmund, UDE and RUB.

Download my resumé.

Primary interests:

Applied Econometrics
Networks
Machine Learning
Data Science

Secondary interests:

Causal Inference
Applied Microeconomics
Social Media

Education

B.Sc. in Economics (with a Big Data Specialization), 2019
ESPOL
M.Sc. in Econometrics, 2024
TU Dortmund, UDE and RUB

Skills

I have been using R since the middle of my undergraduate degree (2017 - present) and have performed some statistical analysis and regression modeling with this language.

Python

In addition to R, I have used Python to perform many activities such as: data collection from Social Media (Twitter), processing and cleaning data, developing of some classification and regression models using ML algorithms, etc.

Gephi

I am very interested in the Networks field and started to use this tool to perform many Social Network Analysis, such as: clustering, centrality metrics, etc. In addition, I complemented this tool with igraph (Python library) to process the network data, construct networks and other tasks.

Neo4j

I began to work with a graph-database (Neo4j AuraDB) and have learned to perform cypher queries in order to get data and upload new data.

Databricks

I have experience in running notebooks and developing pipelines within this environment.

Git

I have started using version control for personal and corporate projects.

Experience

Data Scientist

Maven Road ~ Nodel

Apr 2021 – Dec 2022 Guayaquil, Ecuador

Responsibilities include:

Performed statistical and Time Series Analysis.
Developed Econometrics/Statistical & Machine Learning models.
Detected clusters in Vaping Customers Networks.
Prepared technical reports.

Research Assistant

ESPOL - NC State University

Dec 2021 – Feb 2022 Guayaquil, Ecuador

Responsibilities include:

Predicted some chemical compounds of cocoa with data from leaves and almonds using Machine Learning algorithms like XGBoost, Random Forest and SVRegression.

Research Assistant

CIEC - ESPOL

Dec 2019 – Feb 2021 Guayaquil, Ecuador

Responsibilities include:

Researcher in projects about Socio-economic Evaluation from ESPOL, Cerveceria Nacional, and navigable cruise tourism in the Galapagos Islands.
Performed statistical analysis and developed econometric models (Causal Inference & Propensity Score Matching Methods)

Research Assistant

ESPOL University

May 2019 – Oct 2019 Guayaquil, Ecuador

Responsibilities include:

Supported in the research works:
- “Efficiency of advertising spending analyzed from Twitter Data of companies in the commercial sector during 2018”
- “Network of Shared Administrators and its Relation to the Financial Performance of Ecuadorian Companies: What is the Effect of Sharing Human Capital?"

Data Analyst

Induglobal S.A

Apr 2018 – May 2018 Guayaquil, Ecuador

Responsibilities include:

Analyzed data from a Social Media advertising campaign.

Teaching Assistant

ESPOL University

May 2017 – Sep 2017 Guayaquil, Ecuador

Responsibilities include:

Teaching assistant of Intermediate Microeconomics: Overall Competitive Balance (consumption and production), Market Power (monopoly and price discrimination).

Accomplishments

Research Paper Competition

Superintendecy of companies Nov 2019

Obtained 1st place. Proposed an econometric model that analyzes the efficiency of Advertising Spending of commercial firms by relating Twitter indicators and Dorfman-Steiner condition. In addition, I received additional recognition for participating as co-author in two award-winning research papers.

Hackathon DataJam

ESPOL-i3lab-Broadnet Dec 2018

Obtained 1st place. Developed models to predict transactional behavior and detect anomalies.

Blogs

Time Series Models

This blog shows the application of some time series methods with an example.

MatchIt Example: Nonparametric Preprocessing for Parametric Causal Inference

This blog shows an example about using MachtIt package with Lalonde data set.

Work in progress

Research Project

ESPOL - NC State University Jan 2022

This research aimed to predict the chemical compounds of cocoa through machine learning algorithms, using data from leaves and almonds. Near-infrared devices were used to determine the values of the chemical compounds, and were considered as ground truths to carry out predictions. In addition, the XGBoost, Random Forest, OLS Regression, and SVRegression algorithms were used in this research. Results were compared with the Principal Component Regression models to evaluate the performance of the approaches.

Personal Research Project

Sep 2021

This research is currently being conducted; it aims to determine which metrics had a significant influence regarding the impact that villages suffer during an earthquake. I used socioeconomic variables, such as mobile coverage, education, poverty level, and migration rate. I also estimated the resiliency and concentration indicator using Call Detail Records (CDR) data. Finally, I propose developing an econometric model that associates the number of affected households with the socioeconomic and CDR variables.