Customers

About

Electricity

Legal

COzero News


Articles tagged 'data-science'

A Standard Work-Breakdown Structure for Data Science

A Standard Work-Breakdown Structure for Data Science

In a previous post we introduced how we are approaching data science at COzero, making sure it is an integral part of our DNA as a company. However, data science processes are usually described by some rendition of elements of the scientific method or at times borrowing terms and procedures from statistics. Here we present our interpretation of that process in more detail, which we don't claim is the whole truth but a conscious effort to remove the hype to make the process sustainable from a business perspective.
Continue reading
Data Science in a Business Context

Data Science in a Business Context

Much is said about the successes and failures of the new kid on the block, data science, and arguably there is much hype about it today. But being at the core of COzero's DNA, we made a conscious effort to remove the hype and use data science within our business. What does it mean for a company to do data science? Here is our answer.
Continue reading
Calculating the Characteristic Function

Calculating the Characteristic Function

The Fourier transform of the probability density function is sometimes used to derive certain concepts in Statistics. In this post we describe what it is, its use to describe kernel density estimators and how to calculate it and plot it in Python.
Continue reading
Density Estimation

Density Estimation

Being able to measure the degree of randomness (or of our ignorance) on a bunch of values is a critical tool for data science. In this post we explain the basics and how we are applying this to grab early insights from electricity consumption data.
Continue reading
Hours of Operation and the Gibb's Phenomenon

Hours of Operation and the Gibb's Phenomenon

There are many ways to extract insights out of time series data. In this series of posts we explore some of the mathematical foundations in dealing with an apparently simple but, we will see, challenging task: based on the power electricity profile of a business, determine its hours of operation.
Continue reading