A blog designed like a scientific paper

Real-Time Fraud Detection

Aug. 1, 2018

Miguel González-Fierro

Fraud detection is one of the top priorities for banks and financial institutions, which can be addressed using machine learning. In this post, and in the accompanying code repository, we show how to build a real-time fraud detection system using a boosted tree algorithm and websockets. We also show how to visualize the fraudulent transactions in a map.

A Beginner's Guide to Python Testing

July 24, 2018

Miguel González-Fierro

Test development is key for most software projects. In this post, we are going to discuss 4 different tests: unit tests, smoke tests, integration tests and utility tests. In simple words, unit tests make sure that each class or function behaves as it should, smoke tests make sure that the system works, integration tests make sure that the program results are acceptable and utility tests give an example on how to use a class or function. We will show how to work with these tests in Python.

Demystifying WebSockets for Real-Time Web Communication

July 12, 2018

Miguel González-Fierro

WebSocket is a protocol for bidirectional communication (client-server and server-client), which is part of the HTML 5 API. In simple terms, it enables the interaction between a web client and a web server with lower overheads, facilitating data transfer in real-time. In this post, we will explain what are WebSockets, what is its relationship with older technologies like AJAX and REST APIs and we will provide some resources to help you implement this technology.

Top Mistakes and Learnings from my Startup: Postmortem Samsamia

May 18, 2018

Miguel González-Fierro

Samsamia Technologies was a new communication channel between fashion brands and customers. We developed a search engine for fashion products that allows a smartphone user to find the piece of clothing he or she desires just by taking a picture and buy it instantaneously. In this postmortem, I explain how I built the company, what were the mistakes I made that led to its collapse and what are the learnings I got from it.

10 Ethical Issues of Artificial Intelligence and Robotics

April 1, 2018

Miguel González-Fierro

AI is one of the technologies that could revolutionize the world, some people call it the electricity of the twenty first century. Researchers and professionals need to be aware of the ethical and social implications this technology poses. We are responsible for making robots and AI systems that help and empower humanity.

Stock Price Prediction with LSTMs

March 20, 2018

Miguel González-Fierro

Do you want to make millions in the stock market using Deep Learning? This post will not answer that question, but it will show how you can use an LSTM to predict stock prices with Keras, which is cool, right?

Introduction to Recommendation Systems with Deep Autoencoders

Jan. 28, 2018

Miguel González-Fierro

A recommendation system seeks to understand the user preferences with the objective of recommending items. In this post, we provide an overview of recommendation system techniques and explain how to use a deep autoencoder to create a recommendation system.

A Gentle Introduction to Transfer Learning for Image Classification

Dec. 18, 2017

Miguel González-Fierro

Transfer Learning is expected to be the next driver of Machine Learning commercial success in Image Classification. Reutilizing deep networks is impacting both research and industry. In this post, we explain what is Transfer Learning and when to use its different strategies. The post is accompanied by code in PyTorch performing experiments in several datasets.

Deep Learning for Entrepreneurs

Sept. 30, 2017

Miguel González-Fierro

AI is considered the biggest business opportunity in the new economy, that is expected to generate $15.7 trillion by 2030, according to PwC. Independently of the accuracy of this prediction (probably PwC didn't use machine learning to predict it), the opportunity is huge. In this post, I will discuss some business problems that can be solved with deep learning and how to monetize them.

10 Reasons Why You Should Be Using Git in Software Projects

June 25, 2017

Miguel González-Fierro

Git is a version control software that helps developers to easily and efficiently contribute to software projects. It is the engine that runs on GitHub, the top open-source software development platform. In this post, we analyze 10 reasons why you should be using it.

10 portfolio projects to get you an AI job

Sciblog - A blog designed like a scientific paper

Miguel González-Fierro

Miguel González-Fierro

Miguel González-Fierro

Miguel González-Fierro

Miguel González-Fierro

Miguel González-Fierro

Do you want to make millions in the stock market using Deep Learning? This post will not answer that question, but it will show how you can use an LSTM to predict stock prices with Keras, which is cool, right?

Miguel González-Fierro

A recommendation system seeks to understand the user preferences with the objective of recommending items. In this post, we provide an overview of recommendation system techniques and explain how to use a deep autoencoder to create a recommendation system.

Miguel González-Fierro

Miguel González-Fierro

Miguel González-Fierro

Git is a version control software that helps developers to easily and efficiently contribute to software projects. It is the engine that runs on GitHub, the top open-source software development platform. In this post, we analyze 10 reasons why you should be using it.