Hey there, I'm Alexis—a deep learning enthusiast who loves delving into the exciting realm of data science. My expertise spans radio resource management and reinforcement learning. I'm dedicated to honing my skills, broadening my horizons, and uncovering new opportunities.

• Career Overview
• One-page Resume
• Contact me

Projects

Deep Reinforcement Learning for resource allocation in wireless networks

This project utilized Deep Reinforcement Learning (DRL) and Transfer Learning (TL) for power allocation in wireless cellular networks through simulation environments using Pytorch. This project resulted in three published refereed articles in top journals from Q1 and Q2 in the computer science area.

Recognition of eye diseases with neural networks

This project involves the development of a convolutional neural network (CNN) for detecting ocular diseases using the ODIR-5K database from the Kaggle platform. The CNN was designed using Python with Keras. The model achieved an accuracy of 89.2% in detecting various ocular diseases such as diabetic retinopathy, cataracts, glaucoma, and age-related macular degeneration.

San Diego home price prediction

This project involved collecting data through web scrapping, cleaning and processing data, and training a regression model to predict the home price based on the apartment's neighborhood, square feet, and the number of beds and baths. Finally, the model was hosted on a web page.

Explainable AI (XAI) for beer brand classification

This project combines a Convolutional Neural Network (CNN) based on the VGG16 architecture with GRAD-CAM, an explainable AI method, to interpret the model's focus for beer brand classification. The CNN is fine-tuned with additional layers, achieving an accuracy of 91.6%. Data preprocessing and augmentation were carried out using libraries such as Keras and Sklearn.

Bellabeat capstone project with R

This capstone project is part of the Google Data Analytics Course. The project consists of answering a business task based on the data provided. The data cleaning and data visualization was performed in R.

Sales analysis with pandas

This project analyzes a Kaggle database of electronic devices. The analysis involved cleaning the data and creating visualizations to answer important business questions.

• Full Portfolio
• About me
• Contact me

Career Overview

Alexis Anzaldo

SKILLS

• Programming: Python (Tensorflow, Dagster, Pytorch, Scikit-Learn, OpenCV, Pillow, Matplotlib, Pandas, Seaborn), SQL, HTML/CSS.

• Tools: Microsoft Office, Power BI, Matlab, Labview.

• Spanish: Native. English: B2.

EXPERIENCE

• Applies analytical techniques for valuable insights through data cleaning, feature engineering, and comprehensive analysis.

• Conducts Research and Development of predictive models and algorithms, optimizing inspection processes and guiding strategic business decisions.

• Showcases feasibility of new ideas and technologies through crafting Proof of Concept (PoC) demos and prototypes.

• Develops and implements computer vision algorithms in product development and inspection processes to enhance functionality and performance.

• Drives process optimization initiatives, improving overall efficiency, and integrates AI systems seamlessly into existing workflows.

• Deploys and manages data pipelines using Dagster, ensuring efficient orchestration and integration of data workflows across various projects.

• Evaluation and report of the final defects of the production line. Preparation, correction, and translation of work instructions for operators.

PROJECTS

Resource allocation in wireless networks through Deep Reinforcement Learning

• Accelerated the learning of the conventional Deep Q-Network model for power allocation in wireless networks by up to 77% and improved the network performance by up to 24.7% by proposing different training strategies with transfer learning. Simulations were performed using Python with Pytorch.

• Conceptualized, analyzed, and wrote three published refereed articles in top journals from Q1 and Q2 in the computer science area.

• Conducted a systematic review methodology and identified the 56 most relevant research works implementing machine learning for resource allocation. Performed data extraction, cleaning, and visualization using Excel.

San Diego home price prediction

• Collected and scraped data using BeautifulSoup and preprocessed it by cleaning, handling missing values, and detecting outliers with Python (Pandas, Numpy, and Matplotlib).

• Trained a regression model and achieved an accuracy score of 83.7% using grid search with scikit-learn in Python.

• Deployed the trained model on a Flask server to make predictions and hosted it on a web page using HTML/CSS.

Recognition of eye diseases with neural networks

• Designed a convolution neural network with 89.2% accuracy for detecting ocular diseases using the ODIR-5K database of the Kaggle platform using Python with Keras.

• Managed and planned teamwork tasks for the preprocessing stage involving image formatting, data cleaning, and data augmentation for unbalanced classes.

Explainable AI (XAI) for beer brand classification

• Implemented GRAD-CAM, an explainable AI method, to interpret the Convolutional Neural Network (CNN) decision-making for beer brand classification.

• Fine-tuned the pre-trained VGG16 CNN architecture with additional layers to achieve an accuracy of 91.6%. Data augmentation, preprocessing, and training were performed using libraries such as Keras and Sklearn.

EDUCATION

Universidad Autónoma de Baja California – Mexicali, Baja California, México.

Universidad Autónoma de Baja California – Mexicali, Baja California, México.

Universidad Autónoma de Baja California – Mexicali, Baja California, México.

CERTIFICATIONS

DeepLearning.AI

IBM

Amazon Web Services (AWS)

Google

Academia Lean Sigma

Udemy

PUBLICATIONS

[Under Review] Intelligence-learning driven resource allocation for B5G Ultra-Dense Networks: A structured literature review

Artificial Intelligence Review

https://doi.org/10.21203/rs.3.rs-2763206/v1

Accelerated resource allocation based on experience retention for B5G networks

Journal of Network and Computer Applications

https://doi.org/10.1016/j.jnca.2023.103593

Experience Replay-based Power Control for sum-rate maximization in Multi-Cell Networks

IEEE Wireless Communications Letters

https://doi.org/10.1109/LWC.2022.3202904

Buffer transference strategy for power control in B5G-ultra-dense wireless cellular networks

Wireless Networks

https://doi.org/10.1007/s11276-022-03087-6

CONFERENCES AND PRESENTATIONS

Interference-Aware Power Control for Spectrum Sharing Massive-IoT Communications

14th International Conference on Ubiquitous Computing and Ambient Intelligence (UCAml)

https://doi.org/10.1007/978-3-031-21333-5_46

Deep Reinforcement Learning for Power control in Multi-tasks Wireless Cellular Networks

2022 IEEE International Mediterranean Conference on Communications and Networking (MeditCom)

https://doi.org/10.1109/meditcom55741.2022.9928617

Training Effect on AI-based Resource Allocation in small-cell networks

2021 IEEE Latin-American Conference on Communications (LATINCOM)

https://doi.org/10.1109/LATINCOM53176.2021.9647736

Asignación de Recursos Asistida por IA en Redes Móviles de Nueva Generación

2nd Virtual Seminar on Information Technology

https://www.youtube.com/watch?v=9ly1hSNv318&t=2952s&ab_channel=C%C3%B3digoIA

HONORS & AWARDS

Honorable Mention for PhD Research Project

Universidad Autónoma de Baja California · Oct 2023

Academic Merit for Excellence in PhD Program

Universidad Autónoma de Baja California · Oct 2023

Academic Merit for Excellence in MS Program

Universidad Autónoma de Baja California · Oct 2019