An Overview of Multi-Task Learning in Deep Neural Networks
Note: If you are looking for a review paper, this blog post is also available as an article on arXiv. Table of contents: Introduction Motivation Two MTL methods for Deep Learning Hard parameter sharing Soft parameter sharing Why does MTL work? Implicit data augmentation Attention focusing Eavesdropping Representation bias Regularization... Read more
Why I like the Convolution Theorem
The convolution theorem (or theorems: it has versions that some people would call distinct species and other would describe as mere subspecies) is another almost obviously almost true result, this time about asymptotic efficiency. It’s an asymptotic version of the Cramér–Rao bound. Suppose (hattheta) is an efficient estimator of... Read more
2 Ways to Implement Multinomial Logistic Regression in Python
Logistic regression is one of the most popular supervised classification algorithm. This classification algorithm mostly used for solving binary classification problems. People follow the myth that logistic regression is only useful for the binary classification problems. Which is not true. Logistic regression algorithm can also use to solve the multi-classification problems. So in... Read more
Prophet is Data Science not Statistics, and there is a Difference
Facebook’s prophet forecasting tool illustrates the distinction between a traditional statistical approach compared to the newer machine learning/data science paradigm. This distinction is cultural: it seems that the motivation behind prophet was to quickly make accurate forecasts (predictions), instead of getting bogged down in building models satisfying certain theoretical properties, which may or... Read more
Intro. How I Plan to Teach Myself Deep Learning Using Only Free Resources Learning Deep Learning Series Part 1: Videos Learning Deep Learning Part 2: Online Courses Learning Deep Learning Part 3: Github Repos We’re already familiar with the popularity of big data of the past five or so... Read more
Building a Chatbot: analysis & limitations of modern platforms
The TryoLabs R&D department is in the process of releasing a very interesting AI open source toolkit oriented towards image processing, utilizing state of the art Deep Learning technologies and build upon TensorFlow and Sonnet. Come see Javier Rey, Research Engineer and Alan Descoins, CTO from TryoLabs Speak at ODSC... Read more
Deep Learning Research Review Week 2: Reinforcement Learning
This is the 2nd installment of a new series called Deep Learning Research Review. Every couple weeks or so, I’ll be summarizing and explaining research papers in specific subfields of deep learning. This week focuses on Reinforcement Learning. Last time was Generative Adversarial Networks ICYMI Introduction to Reinforcement Learning... Read more
Artificial Neural Networks (ANN) Introduction
Training a Computer to Recognize your Handwriting Take a look at the picture below above and try to identify what it is. One should be able to tell that it is a giraffe, despite it being strangely fat. We recognize images and objects instantly, even if these images are... Read more
Stacking models for improved predictions
If you have ever competed in a Kaggle competition, you are probably familiar with the use of combining different predictive models for improved accuracy which will creep your score up in the leader board. While it is widely used, there are only a few resources that I am aware of... Read more
Intro to D3
Abstract: Visualizations are the most important way we can intuitively make sense of our data. And the best way to dig deeper into our data is to make those visualizations interactive. Everyone from the New York Times to Google is using interactive visualizations to tell stories and analyze data.... Read more