Folks know that gradient-boosted trees generally perform better than random forest, although there is a price for that: GBT have a few hyperparams to tune, while random forest is practically tuning-free. Let’s look at what the literature says about comparing the two methods.
Numerai is an attempt at a hedge fund crowd-sourcing stock market predictions. It presents a Kaggle-like competition, but with a few welcome twists.
TensorFlow is a new deep learning library from Google. Immediately after release it became the most starred deep learning package on GitHub. From the hype one could conclude that TensorFlow is the best thing since sliced bread. Is it?
Pandas is Python software for data manipulation. We show that some rather simple analytics allow us to attain a reasonable score in an interesting Kaggle competition. While doing that, we look at analogies between Pandas and SQL, a standard in relational databases.
Pedro Domingos’ new book, The Master Algorithm, is a readable overview of machine learning. The author discerns and describes five main schools of thought in the field: symbolists, connectionists, evolutionaries, Bayesians and analogizers. Here’a a piece about how Bayesians fit their models, that is, infer parameter values. Even though the context is Bayes nets, the described method is applicable to almost any model.
If you dig a little, there’s no shortage of recommendation methods. The question is, which model to choose. One of the primary decision factors here is quality of recommendations. You estimate it through validation, and validation for recommender systems might be tricky. There are a few things to consider, including formulation of the task, form of available feedback, and a metric to optimize for. We address these issues and present an example.
The last few weeks have been a time of neural nets generating stuff. By deep nets we mean recurrent and convolutional neural networks, while the stuff is text, music, images and even video.
There is a Kaggle training competition where you attempt to classify text, specifically movie reviews. No other data - this is a perfect opportunity to do some experiments with text classification.
Kaggle has a tutorial for this contest which takes you through the popular bag-of-words approach, and a take at word2vec. The tutorial hardly represents best practices, most certainly to let the competitors improve on it easily. And that’s what we’ll do.
Research into recommender systems took off with the Netflix challenge, which started in 2006. For three years many contenders worked hard to achieve the prescribed error threshold. Finally, in 2009 Netflix awarded the prize, one million dollars.
We can think of two reasons for using distributed machine learning: because you have to (so much data), or because you want to (hoping it will be faster). Only the first reason is good.
Distributed computation generally is hard, because it adds an additional layer of complexity and communication overhead. The ideal case is scaling linearly with the number of nodes; it rarely takes place. Emerging evidence shows that very often, one big machine, or even a laptop, outperforms a cluster.