Adrian Tovar is a Machine learning researcher with 6 year of experience. International publications and conferences. Designed, optimizing and coded neural network algorithms for time series prediction (stock market and weather prediction) and computer vision; as well as doing research on adversarial defenses, information bottleneck principle, time series prediction and recurrent neural networks. With experience in both academy and industry had developed and implemented state of the art algorithms for financial forecasting, costumer behavior analysis, cost optimization, as well as generative models fro increase efficiency.

KEY CONTRIBUTIONS:

• Development and implementation of a forecasting algorithm to estimate stock prices on different countries
• Analysis of client behaviors using machine learning algorithms on big data resulting on an increase of revenue.
• Development of an adversarial defense algorithm, state of the art at the time, to prevent attacks against malicious third parties.

---

CORE COMPETENCIES

• Neural Network Design and Optimization
• Research and Development in Machine Learning
• Application of Machine Learning in Industry and Academia
• Data Analysis and Forecasting
• Problem Solving and Innovation

---

SELECTED PUBLICATION

• Generalization error bounds using Wasserstein distancesGeneralization error bounds using Wasserstein distances, IEEE Xplore. · Jan 17, 2019, Presented at ITW 2018, China.
• A Variational Information Bottleneck Principle for Recurrent Neural Networks, NeurIPS 2021

---

FORMER ROLES

• Data Scientist focusing on implementation fo Machine Learning algorithms to improve revenue
• Machine Learning researcher focusing on developing algorithms for stock market prediction

---

EDUCATIONAL BACKGROUND

• Ph.D in Mathematics, University of Wisconsin – Madison
• B.Sc. in  Mathematics, Unevrsidad Nacional Autonoma de Mexico

---

Renowned for his dedication to advancing cutting-edge machine learning technologies, tailored specifically for dynamic and scalable solutions. His methodology emphasizes the creation of robust, data-driven strategies that significantly boost efficiency and effectiveness in both academic and industrial environments. Through his expert application of neural networks and predictive analytics, Adrian consistently enhances organizational capabilities and delivers insights that propel financial forecasting and customer behavior analysis to new heights.