Currently set to No Index

Facebook built a chatbot to help employees deflect criticism over the holidays

Illustration by Alex Castro / The Verge
Facebook’s public image is in such a disastrous state that the company’s public relations team built an artificial intelligence-powered chatbot to help its employees deflect criticism from family members over the holidays, reports The New York Times.

Read Full Story

So you want to be a data scientist? (Part 1)

“Data Scientist” is definitely the hot new job
description. It is such a new title that the roles and
responsibilities associated with it are still not clearly
defined. What skills do you need to be a data scientist? Well, not
everyone agrees. In this article we will try to be objective by
looking at statistics on what skills employers recruiting data
scientists are looking for.

Read Full Story

Why I’m Not a Fan of R-Squared

The Big Message
People sometimes use (R^2) as their preferred measure of model fit. Unlike quantities such as MSE or MAD, (R^2) is not a function only of model’s errors, its definition contains an implicit model comparison between the model being analyzed and the constant model that uses only the observed mean to make predictions.

Read Full Story

Roulette Wheel Selection Using SQL

Roulette wheel selection is a very useful algorithm found in many applications such as Genetic Algorithm(GA). In GA solutions with higher fitness values are given larger probabilities of being selected to produce children, just like natural evolution. I implemented an Oracle SQL version of the Roulette wheel selection algorithm.

Read Full Story

Random forest interpretation with scikit-learn

In one of my previous posts I discussed how random forests can be turned into a “white box”, such that each prediction is decomposed into a sum of contributions from each feature i.e. (prediction = bias + feature_1 contribution + … + feature_n contribution).I’ve a had quite a few requests for code to do this.

Read Full Story

Zombie semantics spread in the hope of keeping most on the same low road you are comfortable with now: Delaying the hardship of learning better methodology.

Now, everything is connected, but this is not primarily about persistent research misconceptions such as statistical significance.
Instead it is about (inherently) interpretable ML versus (misleading with some nonzero frequency) explanatory ML that I previously blogged on just over a year ago.

Read Full Story

Adding an E-Commerce Admin Dashboard to Saleor

In this tutorial, we’ll explore how to add analytical features, such as reporting and dashboarding to an existing e-commerce web application with Cube.js. We’re going to use Saleor as our e-commerce platform. It is powered by a GraphQL server running on top of Python 3 and a Django 2 framework.

Read Full Story