Jhosimar Arias

Deep Learning Researcher

Biography

I am a part-time lecturer at Universidad Peruana de Ciencias Aplicadas and the founder of ML/DL Meetup AQP, a non-profit community of people interested in artificial intelligence. We organize study groups, symposiums, mentoring sessions, and meetups that emphasize the importance of AI research.

I received my Master's Degree in Computer Science at Institute of Computing , University of Campinas. My thesis focuses on the study of semi-supervised and unsupervised clustering based on deep generative models.

I graduated with a bachelor's degree in Systems Engineering from National University of Saint Agustine. During my undergraduate years I had the opportunity to participate in several competitions such as ACM/ICPC international programming contest representing my university and online competitions (TopCoder, HackerRank, LeetCode).

As part of my professional experience I worked as software engineer for zAgile, a startup located in San Francisco, integrating software technologies such as Salesforce, Confluence and Jira.

Interests

Deep Learning
Computer Vision
Probabilistic Machine Learning

Education

MSc. in Computer Science, 2018

University of Campinas (Brazil)
BSc. in Systems Engineering, 2012

National University of Saint Agustine (Peru)

News

[06/22] - Received a Scholarship to attend the International Conference on Machine Learning (ICML).
[04/22] - Grateful to receive a registration grant to attend the conference on Computer Vision and Pattern Recognition (CVPR).
[04/22] - Accepted and granted a scholarship in the Nordic Probabilistic AI School (ProbAI) School, Helsinki, Finland.
[12/21] - Received a Scholarship to attend the Neural Information Processing Systems Online Conference (NeurIPS).
[10/21] - Received a Scholarship to attend the International Conference on Computer Vision (ICCV).
[08/21] - Received a Scholarship to attend the Conference on Knowledge Discovery and Data Mining (KDD).
[08/21] - Invited to give a talk on the "Limitations and New Frontiers of Deep Learning" for Data Science Woman Peru.
[07/21] - Accepted in the Machine Learning Summer School (MLSS) 2021 Taipei.
[07/21] - Accepted in the Deep Learning + Reinforcement Learning (DLRL) Summer School.
[07/21] - Received a Scholarship to attend the International Conference on Machine Learning (ICML).
[05/21] - Invited to give a talk on Deep Learning at Public Technological Higher Education Institute - "Espinar", Peru.
[05/21] - Grateful to receive a registration grant to attend the conference on Computer Vision and Pattern Recognition (CVPR).
[04/21] - Received a Scholarship to attend the International Conference on Learning Representations (ICLR).
[03/21] - Received a Scholarship Pass to attend the Virtual Open Data Science Conference East (ODSC).
[03/21] - Teacher of "Introduction to Deep Learning" course at Universidad Peruana de Ciencias Aplicadas (UPC), Peru.
[03/21] - Part-time lecturer at Universidad Peruana de Ciencias Aplicadas (UPC), Peru.
[01/21] - Lead organizer of the III Peruvian Symposium on Deep Learning (SPDL).
[12/20] - Received a Scholarship to attend the Neural Information Processing Systems Online Conference (NeurIPS).
[11/20] - Invited to give a talk about "Limitations and New Frontiers of AI" at University of San Martin de Porres, Peru. [recording]
[10/20] - Glad to give a talk about Markov Decision Processes as part of the Reinforcement Learning Study Group. [recording]
[09/20] - Glad to give a talk about Batch Normalization as part of the Deep Learning Study Group. [recording]
[08/20] - Received a Scholarship to attend the Conference on Knowledge Discovery and Data Mining (SIGKDD).
[07/20] - Started the organization of the Deep Learning Study Group (in spanish) based on the Deep Learning Specialization of Coursera.
[07/20] - Received a Scholarship to attend the International Conference on Machine Learning (ICML).
[06/20] - Invited to give a talk about "Deep Learning: Applications, Challenges and Opportunities" at University of San Martin de Porres (USMP), Peru.
[06/20] - Invited to give a talk about Competitive Programming at Technological University of Peru.
[05/20] - Teacher of "Introduction to Machine Learning" course offered by La Salle University, Peru.
[04/20] - Received a Scholarship to attend the International Conference on Learning Representations (ICLR).
[04/20] - Received a Scholarship Pass to attend the Virtual Open Data Science Conference East 2020.
[02/20] - Co-lecturer of the Machine Learning course offered in the "Data Science Certificate Program" by La Salle University in Arequipa, Peru.
[01/20] - Presented a practical session about "Deep Learning Fundamentals" at the "Second Peruvian Symposium on Deep Learning" in Arequipa, Peru.
[01/20] - Lead organizer of the Second Peruvian Symposium on Deep Learning in Arequipa, Peru.

Past News

[12/19] - Invited to give a talk about "The Power and Limits of Deep Learning" at "Chapter Week - Systems Engineering & Informatics" in Arequipa, Peru.
[10/19] - My work "Deep Clustering using MMD Variational Autoencoder and Traditional Clustering Algorithms" was accepted in the workshop of Sets & Partitions (NeurIPS), Vancouver, Canada.
[10/19] - My work "Semi-supervised Learning using Deep Generative Models and Auxiliary Tasks" was accepted in the 4th Bayesian Deep Learning workshop (NeurIPS), Vancouver, Canada.
[09/19] - Invited to give a talk about "Conferences and Opportunities in Artificial Intelligence" at "Artificial Intelligence Seminar" in Arequipa, Peru.
[06/19] - Attended the Conference on Computer Vision and Pattern Recognition (CVPR), Long Beach, CA.
[06/19] - Presented my work of semi-supervised classification as oral presentation and poster at the Latinx in AI Workshop (ICML), Long Beach, CA.
[06/19] - Invited to give a talk about "Foundations and Applications of Neural Networks" and to be part of the tech expo with my community at Xpotron 2019: Control of Dynamical and Aerospace Systems in Arequipa, Peru.
[05/19] - Invited to give a talk about "The Power and Limits of Deep Learning" and a practical session about "Unsupervised Learning" at Computer Day Puno 2019: Artificial Intelligence in Puno, Peru. [slides] [code]
[02/19] - Invited to give a talk about Deep Learning at the Machine Learning Seminar in Cuzco, Peru. [slides]
[01/19] - Presented my work of unsupervised clustering at the "First Peruvian Symposium on Deep Learning" in Arequipa, Peru. [slides]
[01/19] - Presented a tutorial about the Foundations of Unsupervised Deep Learning at the "First Peruvian Symposium on Deep Learning" in Arequipa, Peru. [slides]
[01/19] - Lead organizer of the First Peruvian Symposium on Deep Learning in Arequipa, Peru.
[01/19] - Presented my work of semi-supervised clustering at the X Peruvian Symposium on Artificial Intelligence in Arequipa, Peru. [slides]
[05/18] - Founded the community ML/DL Meetup AQP in Arequipa, Peru.
[02/18] - Defended my master's thesis at UNICAMP in Campinas, Brazil. [slides]
[01/18] - Invited to give a talk about my work presented at NIPS. International Conference on Computer Research in Arequipa, Peru. [slides]

Publications

Deep Clustering using MMD Variational Autoencoder and Traditional Clustering Algorithms

Jhosimar Arias

Sets & Partitions Workshop (NeurIPS 2019)

PDF

Semi-supervised Learning using Deep Generative Models and Auxiliary Tasks

Jhosimar Arias

4th workshop on Bayesian Deep Learning (NeurIPS 2019)

PDF

Deep Generative Models for Clustering: A Semi-supervised and Unsupervised Approach

Jhosimar Arias

Master Thesis

PDF Slides

Is Simple Better?: Revisiting Simple Generative Models for Unsupervised Clustering

Jhosimar Arias and Adín Ramírez

Second workshop on Bayesian Deep Learning (NIPS 2017)

PDF Code

Learning to Cluster with Auxiliary Tasks: A Semi-Supervised Approach

Jhosimar Arias and Adín Ramírez

30th Conference on Graphics, Patterns and Images (SIBGRAPI 2017)

PDF IEEE Code Slides

Projects

Few-Shot Image Classification

Implemented the few-shot learning assigments of the CS330 course offered by Stanford University, which include MANN, MAML and ProtoNet. Worked with Python and Tensorflow.

GMVAE for clustering

In this project, I implemented a Gaussian Mixture Variational Autoencoder by representing the Categorical latent variable with the Gumbel-Softmax distribution avoiding the problem of multiple gradient estimators used in marginalization. Experiments showed around ~80% of clustering accuracy with multilayer perceptrons. Worked with PyTorch and Tensorflow.

Code

CS231n: Convolutional Neural Networks for Visual Recognition

Implemented the assignments given by the CS231n course offered by Stanford University, which covers different machine learning topics including image classifiers (kNN, SVM, Softmax), CNNs, RNNs, LSTMs and GANs. Worked with python and Tensorflow.

Code

Trademark Image Retrieval using Deep Feature Maps

In this work, I present a study of transfer learning applied to trademark image retrieval. Initially selective search is used to obtain region proposals, which are forwarded through a pretrained CNN architecture (AlexNet, GoogleNet and ResNet) on the ImageNet dataset. Feature representations are improved by developing feature aggregation methods (avg-pool, max-pool and R-MAC) over intermediate layers. Finally re-ranking based on graph query specific fusion algorithm was applied to improve the results. Experiments demostrate that intermediate layers produce better results for image retrieval. It was possible to increase in ~15% the baseline (features of last layers) mean average precision (mAP). Worked with python and Caffe.

Slides

See all projects

Leadership

Machine Learning & Deep Learning Meetup Arequipa (ML/DL Meetup AQP)

Jhosimar Arias (Founder, Instructor and Lead Organizer)

ML/DL Meetup AQP is a Non-Profit community of people interested in Artificial Intelligence (AI), particularly in Machine Learning (ML) and Deep Learning (DL). We organize meetups in person and online reviewing books, lectures, research papers, and courses from top Universities and MOOC's. It's an open community to people of all levels of knowledge.

Among the activities we have organized:

Study groups on Machine Learning, Deep Learning and Reinforcement Learning
"III Peruvian Symposium on Deep Learning - SPDL 2021"
"II Peruvian Symposium on Deep Learning - SPDL 2020"
"I Peruvian Symposium on Deep Learning - SPDL 2019"
Meetups and AI seminars at different universities in Arequipa
Paper discussions on relevant Deep Learning topics
Talks by invited speakers

Talks

Introduction to Deep Learning

Jun 20, 2021 Public Technological Higher Education Institute - "Espinar", Peru

I Training Course in Computer Science for the Programming and Development of Multiplatform Systems

Recording

Limitations and New Frontiers of AI

Nov 20, 2020 University of San Martin de Porres (USMP), Peru

Artificial Intelligence: a scenario for Digital and Technological Transformation (Virtual Webinar)

Recording

Markov Decision Processes

Oct 01, 2020 ML/DL Meetup AQP, Peru

Reinforcement Learning Study Group

Recording

Hyperparameter Tuning, Batch Normalization and Multiclass Classification

Sep 06, 2020 ML/DL Meetup AQP, Peru

Deep Learning Study Group

Recording

Deep Learning: Applications, Challenges and Opportunities

Jun 19, 2020 University of San Martin de Porres (USMP), Peru

Information technology trends and best practices for e-learning and business continuity in times of crisis (Virtual Webinar)

Competitive Programming

Jun 09, 2020 Technological University of Peru (UTP), Peru

Media

Oral Presentation in LXAI at ICML 2019

Jun 10, 2019 LXAI @ ICML, Long Beach

LatinX in AI Workshop at ICML 2019

Recording Slides

See all talks

Teaching

2022-2

CC126 - Introduction to Algorithms, Universidad Peruana de Ciencias Aplicadas (UPC), Peru
CC227 - Introduction to Deep Learning, Universidad Peruana de Ciencias Aplicadas (UPC), Peru

2022-1

CC100 - Programming I, Universidad Peruana de Ciencias Aplicadas (UPC), Peru
CC68 - Algorithms and Data Structures, Universidad Peruana de Ciencias Aplicadas (UPC), Peru

2021

CC199 - Emerging Topics in Technology (Introduction to Deep Learning), Universidad Peruana de Ciencias Aplicadas (UPC), Peru

2020

Introduction to Machine Learning, Short Course, La Salle University, Peru
Machine Learning (Co-Instructor), Data Science Certificate Program, La Salle University

2016

Competitive Programming, Short Course, CITEC, Peru

Blog

Algorithms and More

In 2012, I started a blog in Spanish on algorithms and programming in order to help people better understand the algorithms and data structures used in competitive programming and programming projects. Although I am not very active on this blog, the algorithm explanations I posted are very useful nowadays (more than ~12000 views per month).

The explanations are given step by step with graphical examples, code and exercises. Some of the most interesting posts based on user statistics are:

Jhosimar Arias

Deep Learning Researcher

Biography

Interests

Education

News

Publications

Deep Clustering using MMD Variational Autoencoder and Traditional Clustering Algorithms

Semi-supervised Learning using Deep Generative Models and Auxiliary Tasks

Deep Generative Models for Clustering: A Semi-supervised and Unsupervised Approach

Is Simple Better?: Revisiting Simple Generative Models for Unsupervised Clustering

Learning to Cluster with Auxiliary Tasks: A Semi-Supervised Approach

Projects

Few-Shot Image Classification

GMVAE for clustering

CS231n: Convolutional Neural Networks for Visual Recognition

Trademark Image Retrieval using Deep Feature Maps

Leadership

Machine Learning & Deep Learning Meetup Arequipa (ML/DL Meetup AQP)

Talks

Introduction to Deep Learning

Limitations and New Frontiers of AI

Markov Decision Processes

Hyperparameter Tuning, Batch Normalization and Multiclass Classification

Deep Learning: Applications, Challenges and Opportunities

Competitive Programming

Oral Presentation in LXAI at ICML 2019

Teaching

2022-2

2022-1

2021

2020

2016

Blog

Algorithms and More

Contact