henryzord / mimic

Implementation by Henry Cagnini for the MIMIC algorithm described in: De Bonet, Jeremy S., Charles L. Isbell, and Paul Viola. "MIMIC: Finding optima by estimating probability densities." Advances in neural information processing systems (1997): 424-430.

👋 Hi!

My name is Henry Cagnini. I have a PhD in Computer Science by Pontifical Catholic University of Rio Grande do Sul (PUCRS), with a six-month internship at University of Kent. While I also got my Master's in Computer Science at PUCRS, I have a bachelor's degree in Computer Science by Federal University of Santa Maria (UFSM).

I was born and raised in Santa Maria, Rio Grande do Sul, Brazil.

💽 What do you do?

I'm currently an IT Analyst at Federal University of Santa Maria (UFSM). I mostly work with business intelligence, data mining, and machine learning.

I was also...

• A Substitute Teacher at Industrial Technical College of Santa Maria (CTISM) (High School and Undergratuate);
• A teacher at various institutions, such as PUCRS, Uniritter, Feevale, TargetTrust (Postgraduate);
• An intern at Motorola Mobility Inc;
• An intern at Enovative Design and Technology.

You can check some of my current and previous work at these organizations:

• https://github.com/COPLIN-UFSM
• https://github.com/CTISM-Prof-Henry

Research interests

My research interests are evolutionary algorithms, machine learning, and the teaching of computer science topics.

🎨 What are your hobbies?

I like to...

• Play videogames 🎮
• Travel ✈️
• Eat at cafés ☕
• Be with the people I love 💑

🌐 Social networks

• Twitter
• Google Scholar
• dblp
• Lattes CV (in PT-BR)

📫 How to reach me

E-mail me: [email protected]

💻 Competences

My competences sorted by the most to least competent are*:

Status	Programming languages	Marking languages	Technologies and topics
5️⭐	Python, Java, C	LaTeX, Markdown	Evolutionary Algorithms, Machine Learning, Statistics, Data visualization
4️⭐	C++	HTML, RestructuredText
3️⭐	Assembly, Android, Javascript	CSS	Frontend, Backend, Databases
2️⭐	OpenCL, CUDA, Prolog, Haskell, PHP		Deep Learning
⭐	You're not here for my one-star competences, are you?

* Non-extensive list. Just the ones from the top of my head

⚡ Fun fact

Did you know the humanity already solved one of the seven Millenium Problems?

Metrics

mimic's Issues

Fix sampling function

The PMF in the sampling function is wrong. Numpy.random.choice is setting the probability to draw a value based on the size of the sample, instead of drawing n samples with the p probability for each drawing. Fix this.